LETTER FROM HENRY BIENEN TO ROBERT M. GATES
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
CIA-RDP90G01353R001700010025-0
Release Decision:
RIPPUB
Original Classification:
K
Document Page Count:
90
Document Creation Date:
December 23, 2016
Document Release Date:
January 14, 2013
Sequence Number:
25
Case Number:
Publication Date:
April 15, 1988
Content Type:
MEMO
File:
Attachment | Size |
---|---|
CIA-RDP90G01353R001700010025-0.pdf | 4.79 MB |
Body:
1
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14:
CIA-RDP90G01353R001700010025-0
#43
0 Next 3 Page(s) In Document Denied
ifil)bliP4
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14:
CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
1
PrinC.CtO1144111VerSityCenter-ofintektiiiiiiiiirSiridres
Corwin Hall
- Princeton, New Jersey 08544
Tel: (609) 452-4851/6404
April 8, 1988
Mr. R ert M. Gates
Dep y Director
T Central Intelligence Agency
ashington, D.C. 20505
Dear Mr. Gates:
1 1-1.51.uino 11.1.6.4...,
887138Ix/2
Henry S. Biencn, Director
James S. McDonnell Distinguished
University Professor
STAT
On behalf of the students in the policy conference, I want to thank you
very-much--for-your-detailed-comments-and-criticisms, as-well-as-yout-praise
for parts of the conference Report.
I hope you realize that this Report does not reflect either my views nor
my understandings. I, too, thought that parts of the Report on covert
activities were not always accurate. Indeed, I have differences with parts
on the producer-consumer dexus too. My role was to provide a context for the
conference, provide information and bibliography where I could, and arrange
the participation of key policy makers like yourself. I did not want to
impose my own views or analysis on the student conferees.
We are going to publish the conference Report as an occasional paper of
the Center of International Studi--I-Woill-dry much like to also publish
your letter as a comment on the Report if you agree. I ask your permission
to do so. Perhaps someone from your office can call me or perhaps
can do so.
STAT
I enclose a comment from one of my colleagues on the report. Andy
Marshall also called me with his comments.
Again, many thanks for your comments and your efforts which got us off
to a great start. I am glad that you were able to put your talk to us before
a wide audience.
HB/gak
Enc.
Sincerely,
Henry Bienen
7,11
WOODROW WILSON SCHOOL OF PUBLIC AND INTERNATIONAL AFFAIRS
LaCi
EXEC
REG
STAT
L.-261-qt-
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
STAT
R
Next 2 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
ER 1381/1-88
The Deputy Director of Central Intelligence
Washington. D. C.20505
April 4, 1988
Professor Henry Bienen
Center of International Studies
Princeton University
Corwin Hall
Princeton, New Jersey 08544
Dear Professor Bienen:
Many thanks for sending me the final report for the Policy
Conference on Intelligence. I'm impressed by the quality of
the work and the sophistication of the analysis. I found the
sections on the intelligence agencies, problems inherent in the
nature of analysis and the producer/consumer nexus all to be
insightful and well done. While there obviously are specifics
with which I might disagree or have a different point of view,
the analysis itself is as good as any I have read. Further,
the recommendations are, for the most part, sensible and worthy
of serious consideration.
I found the section on covert action, however, to be more
uneven with a number of factual errors, some outdated
information, and acceptance of newspaper allegations as fact
A few examples are worth noting:
There is a suggestion on page 49 that the Deputy
Director of Operations, the DDCI and the DCI should
sign off on all covert actions as well as on periodic
updates. Actually, for a number of years, those three
officers have been required to sign off on all covert
action findings as well as periodic updates. Beyond
this, there are regular evaluations of these operations
which are provided to these three officials and other
senior managers.
The paper says on page 53 that, "Since 1985 it has been
known that Contras and Contra supporters are heavily
involved in the US narcotics trade in order to earn
cash to purchase weapons." At minimum, that is an
unproven allegation; such a thing is not "known" at
31
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
this time. A number of investigations are under way to
ascertain the validity of such allegations. We are
cooperating fully in these investigations. In any
event, such a sweeping assertion is unjustified, even
on the basis of the allegations that have been made.
(In this connection, the next sentence notes that, "In
1985 the AP carried a story on the largest cocaine
seizure in West Coast history; the convicted smuggler
said that the profits were meant to buy Contras arms."
I would think that some skepticism might be expressed
when quoting a convicted drug smuggler trying to
justify or excuse his actions.)
On page 55 the proposal is made to have covert actions
submitted to the DDI and the State Department for
review. Again, with the single exception of the Iran
Findings, the DDI and State have reviewed all proposed
covert action findings for at least the last seven
years. And, as you may have read, DCI Webster has
established new approval procedures for all Covert
Actions that ensure the participation not only of the
DDI but of other non-DO Agency officials such as the
General Counsel, Director of Congressional Affairs,
Comptroller, etc.
On page 59 a recommendation warns the policymaker to be
"extremely wary of employing covert action, for the
future ramifications of large scale military secret
operations can and often are extremely serious. Covert
action is not practical ...." Experience would suggest
that there are, indeed, often considerable risks
attendant to paramilitary covert actions. However,
there are a number of other types of covert action
which traditionally have had broad bipartisan support
on the Hill, have not leaked, and have been effective.
Differentiating between the risks attendant to
paramilitary covert actions and others would be a more
accurate reflection of past experience.
The paper asserts on page 64 that Casey prevented the
IOB from fulfilling its mission by moving the CIA's
General Counsel to a new office four blocks away from
CIA Headquarters at Langley, thereby isolating it from
many of his sources of information. In fact, the
decision to move the General Counsel's office out of
CIA Headquarters was made purely on the grounds that
growth in the analytical and operations directorates
whose officers need access to one another and senior
agency managers -- required making available more space
at Headquarters. It was CIA's senior professional
management that recommended to Casey that the Office of
General Counsel be temporarily relocated until our new
2
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14 : CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Headquarters addition is complete. The decision had
absolutely nothing to do with the IOB or attempting to
isolate the General Counsel from his sources of
information. Indeed, Casey was reluctant to make the
decision and did so only under considerable pressure
from senior managers. (DCI Webster has determined that
the General Counsel's office should be returned to CIA
Headquarters as quickly as possible, and they are now
scheduled to move into the new building complex next
January. In the interim, the current General Counsel
maintains his primary office at our Headquarters, as do
additional lawyers who work closely with specific
components of the four directorates.)
With respect to the recommendations for the Inspector
General on page 65, the recommendations were
implemented some weeks ago by DCI Webster -- including
rank equal to that of the Agency's deputy directors.
The overall role and activities of the office are being
strengthened along with new powers for investigation.
Moreover, the IG can and does take matters to the I0B,
just as the General Counsel brings matters to the
attention of both the IOB and the Justice Department.
Concerning the alleged crippling of CIA's interaction
with academic resources and American business (on page
69), this is a vast overstatement. American business
has never flagged in its willingness to talk to our
people, volunteer information, and contribute to our
analytical efforts. Moreover, I think it is fair to
say this Agency now has more contacts than ever with
universities on substantive issues, for recruitment and
for consulting. It is indeed quite rare for
academicians to turn down the opportunity to talk with
us, particularlY on analytical issues. Moreover, we
have strong support for these contacts from university
administrations across the entire country. Along these
lines, I would cite your own invitations to CIA
officers to participate in your conference, as well as
the recent contract signed with the Kennedy School at
Harvard University to provide training and education
opportunities for our officers.
Finally, from a purely personal standpoint, the
description on page 50 of the Agency and the
Iran/Contra affair is wrong in important particulars.
Setting aside Mr. Casey's role, which is difficult to
clarify because of his illness and death, neither I nor
my predecessor were aware of any improper activities
with respect to the Contras during the period when CIA
support was severely restricted. Any discussion of the
"Iran-Contra" affair must break it down into its
3
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14 : CIA-RDP90G01353R001700010025-0
component constituent elements: proscribed support for
the Contras in which this Agency was not involved (with
the possible exception of one or two officers operating
in violation of our rules), and the Iran affair, in
which no allegation has been made of illegal or
improper activity by the Agency -- however unwise the
policy may have been. Furthermore, the lunch with
North from my perspective was specifically to inquire
whether he was aware of any CIA involvement in the
private benefactor activity and in the crash of the
plane that led to the capture of Eugene Hasenfus. In
short, the discussion on page 50 is quite muddled and
inaccurate.
As I said at the outset, I found the report in general to
be of very high quality and most impressive. I mention the
specific points above -- the most important problems, though
not an exhaustive recitation -- only in the event that one or
another of the participants might consider publishing the
report in part or in its entirety and knowing that, as good
scholars, they would want to ensure that it is as accurate as
possible.
Again, I very much welcomed the opportunity personally to
be involved in the Conference. I found the students
extraordinarily well informed and would venture that there is
nothing comparable to the work of the Conference on this
subject that is as sophisticated as what I observed and now
have read.
Congratulations on running a first-class program.
DDCl/RMGates/de
DISTRIBUTION:
0 - Addressee
1 - D/PAO w/incoming
- FOIA, w/incoming
ER, w/incoming
- DDCI Chrono, wo/incoming
1 - DDCI Subject File, w/incoming
4
STAT
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
j Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
1 ER 1381X-88
Princeton University
Center of International Studies
Corwin Hall
Princeton, New Jersey 08544
Tel: (609) 452-4851/6404
March 2, 1988
Mr. Robert Gates
Deputy Director
Central Intelligence Agency
Washington, D.C. 20505
Dear Mr. Gates:
Henry S. Bienen, Director
James S. McDonnell Distinguished
University Professor
Here is the final conference report for our Policy Conference on
Intelligence. Thank you again for all your help.
Sincerely,
Henry Bienen
/6412-juiv
HB/gak
Enc.
WOODROW WILSON SCHOOL OF PUBLIC AND INTPRIVATTCMIAT A Pr ATD c
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
47!
The United
States Intelligence Communityc,
Conference Report
Henry S. Bienen, Conference Director
Andrew Balson
Marshall Scott Huebner
Senior Commissioners
The Woodrow Wilson School of Public and International Affairs,
Princeton University
Fall 1987
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
Table of Contents
Introduction and List of Abbreviations
Organizational Problems
1
Agency Specific Issues
6
The Central Intelligence Agency
8
The Defense Intelligence Agency
13
The Bureau of Intelligence and Research
23
Problems Inherent in the Nature of Analysis
28
Consensus Building vs. Competitive Analysis
28
Political Surprise
29
Problems on the Level of the Individual Analyst
31
Evaluation of Analysis
34
Personnel Quality
36
The Producer/Consumer Nexus
40
Covert Action
46
Long Term and Regional Issues
56
Oversight
62
Executive Oversight
'62
Congressional Oversight.
66
Public Opinion and the Intelligence Community
68
Summary Of Recommendations
72
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
The Woodrow Wilson School Policy Conference on "The Role of Intelligence in U.S.
Foreign Policy" convened during the fall semester of 1987 to examine issues of the intelligence
community and its role in the foreign policymaking process. In the course of its research, the
conference met with a wide variety of speakers from many areas and levels of the intelligence
community. We would like to thank them both for their candor and for the invariably well
prepared and illuminating presentations that they gave.
The controversial nature of the United States intelligence community creates difficulties for
any group attempting to examine it. The issues involved, in addition to being remarkably
complex, are also rife with far-reaching political and ethical implications. The solutions are
often personal and imperfect, making the production of a critical and objective paper on the
various aspects of the intelligence community nearly impossible. With that caveat in mind, the
Woodrow Wilson School Policy Conference on "The Role of Intelligence in U.S. Foreign
Policy" has produced what we intend to be a critical and unbiased look at the United States
intelligence community, in the hope that we might contribute some insights to the debate over
the community's problems and possibilities.
Our observation of the United States intelligence community revealed certain general
tendencies. Most striking is the importance of the high degree of informal contact that occurs
within and between the intelligence agencies and between the intelligence and policymaking
bureaucracies. Personal contact significantly eases bureaucratic impediments to efficiency at all
levels of the system. Many of our recommendations draw heavily on this observation. Our
examination has led us to eschew drastic structural reorganization (with one prominent
exception) in favor of reforms geared to promoting informal and relatively unstructured
mechanisms to improve the system as it presently exists. We have found that the structures
now in place, both for the production and the oversight of intelligence, are generally sound.
Many of the problems in the system are the result of attitudinal shortcomings and
misconceptions among analysts and policymakers, rather than structural deficiencies.
L_ Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Only after having spent substantial time wrestling with the issues of the intelligence
community did the conference come to its second major realization -- that the activities of the
intelligence community are rarely, if ever, ends unto themselves -- but rather are useful only
through the contributions they make to policy. Though the point seems evident in retrospect, it
is easy to lose this perspective when immersed in the subject matter. This observation bears on
nearly all of the issues considered. For instance, many of the problems in the dissemination of
the analytical product are problems with consumers, not with those who produce the
intelligence work. Similarly, the success or failure of covert operations may have little to do .
with the nature of the operations themselves and everything to do with the policy environment
in which they occur. It is a serious concern of the intelligence producers that their work is
often ignored or unread by the policymakers whom they serve, one which we address at some
length in the body of the report.
Our third general observation, related to the first two, concerns the role of the human
element in the intelligence community. Critiques often disregard the essential characteristic of
the intelligence bureaucracy -- that it is comprised of fallible, human components -- in a search
for perfect solutions to intelligence problems. It is essential to recognize that we will never
create a perfect intelligence community. What we must do is encourage personal and attitudinal
improvements among those who comprise the intelligence and policymaking bureaucracies.
Many of our recommendations are intended to affect the human side of intelligence, and
concern the selection, training, promotion and evaluation of personnel.
Worth singling out among our observations of the human influences on the functioning of
the intelligence bureaucracy is the importance of attitudinal change. Many of the problems
across the spectrum of intelligence issues could at least be made less severe by raising
community consciousness of their existence. When confronted with problems of mindset in
analysis, politicization in the producer/consumer interface, or the potentially harmful effects of
certain covert operations, members of the community are often unwilling to admit or simply
II
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
unaware of their own flaws. Convincing intelligence analysts, managers, and operators of
their shortcomings is made particularly difficult by the fact that they usually do high quality
work. Attitudinal changes will not eliminate errors. Increased consciousness will, however,
reduce their likelihood, make their consequences less severe, and engender quicker responses
to mistakes when they do occur. Our belief in the importance of making explicit recurrent
problems in intelligence leads us to propose several recommendations specifically tailored
towards raising consciousness within the intelligence community.
Another important problematic issue within the community are inadequate levels
communication. In response, other of our recommendations are designed to increase
interaction both in relations within the community itself and between the producers and the
consumers of intelligence. It is especially important to increase the communication between
working level intelligence staffers and both their junior and senior level counterparts in the
policy arena -- too often, the expertise and knowledge of the analytical staff are underutilizied
because it is their superiors alone who have occasional access to the consumers of intelligence.
In short, this report gives a broad overview of the intelligence community and attempts to
pinpoint both the areas that are most directly in need of improvement and the solutions that
seem most likely to bring it about. For more detailed discussions of the issues examined below
see the fifteen Junior Independent Papers of the Woodrow Wilson School Policy Conference
on "The Role of Intelligence in Foreign Affairs" which formed the basis for this report
II
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Conference Director: Henry S. Bienen, Director of the Center of International Studies,
Princeton University.
Senior Commissioners: Andrew B. Balson
Marshall Scott Huebner
Conference Members:
Michael Balaoing Wayne McKinzie
Andrew Blake John Meier Ill
Laura Coolidge Scott Perekslis
Abigail Franklin Henry Ricardo
Jonathan Jaffee Alinor Sterling
James Kirby Robin Traylor
Clifford Levy Lori Zeltser
David Littel
IV
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
List of Abbreviations
CIA- Central Intelligence Agency
CS- Civil Service
DAS- Defense Attache System
DOD- Department of Defense
DCI- Director of Central Intelligence
DDCI- Deputy Director of Central Intelligence
DDIA-Director of the Defense Intelligence Agency
DIA- Defense Intelligence Agency
DIC- Defense Intelligence College
DE- Directorate of Estimates
DNI- Director of National Intelligence
FSO- Foreign Service Officer
FY- Fiscal Year
I0B- Intelligence Oversight Board
INR- The Bureau of Intelligence and Research
JCS- Joint Chiefs of Staff
MIO- Military Intelligence Officer
MIS- Military Intelligence Service
NFIB-National Foreign Intelligence Board
NIO- National Intelligence Officer
NIE- National Intelligence Estimate
NSA- National Security Agency
NSC- National Security Council
OPC- Office of Policy Coordination
OSD- Office of the Secretary of Defense
OSO- Office of Strategic Operations
PFIAB-President's Foreign Intelligence Advisory Board
SECDEF- The Secretary of Defense
SIG-I-Senior Intelligence Group for Intelligence
SIS- Strategic Intelligence Service (GB)
SOE- Special Operations Executive (GB)
SSG- Special Situations Group
V
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
,o
Organizational Problems
Certain organizational structures, while necessary for the proper functioning of the intelligence
agencies, have side effects that may reduce the quality of the analytical product In particular, the
hierarchical nature of the intelligence community, the specialization required of intelligence analysts,
and the degree of centralization needed to adequately control the various elements of the community
can all serve to undermine its analytical purpose. It is of paramount importance, therefore, to find
ways to limit the damage to the efficient operation of the intelligence bureaucracies caused by the
necessities of organizational structure.
Because the structure of the hierarchy determines rank and status, it can hamper performance by
impeding information flow through the bureaucracy. Subordinates will feel pressure not to report
data that conflicts with the established opinions or routines of their supervisors because they fear
resulting penalties. Similarly, they might not relay information which puts their own performance
in a bad light, because it could effect their position in the hierarchy.1
Within the intelligence community these constraints manifest themselves particularly in the lack
of contact between the upper, middle, and lower levels of the intelligence hierarchy. The problem
is exacerbated by the fact that upper level intelligence officials often occupy their positions for a
short time and thus do not develop sufficient informal ties with lower level analysts to overcome
bureaucratic limitations. These organizational restrictions on information flow contribute to
insufficient levels of upper echelon expertise within specific subject areas. Communication between
various levels in the intelligence community hierarchies must be facilitated in order to allow
supervisors to better benefit from the expertise of their subordinates.2 In order to counteract
hierarchical constraints on analysis and thereby enhance communication within the intelligence
bureaucracies, we underscore the following:
1. Further enhancement of informal modes of communication, rather than a more formal
restructuring of the hierarchy, will most productively reduce impediments to information flow
within the intelligence community.
1Glenn P. Hastedt, "Organizational Foundations of Intelligence Failures," in A. Maurer, M. Tunstall, and J.
Keagle eds., Intelligence Policy and Process (Boulder: Westview Press,1985), p.141.
2Harold Wilensky, Organizational Intelligence (New York: Basic Books, 1967), pp.41-62.
1
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
To counteract hierarchical constraints, members of the analytical branches of the intelligence
community tend to develop informal ties with their peers throughout the intelligence community.
Informal tasking and information sharing already plays an important part in the analytical process.
These contacts, that allow analysts to bypass bureaucratic constraints, contribute markedly to the
dissemination of information and analysis throughout the community and should be encouraged. In
addition, increased informal contact between upper and lower levels within the community would
improve the flow of information upward through the hierarchy. Communication could be facilitated
by weekly or biweekly briefings of top officials in the intelligence community by subordinates on
certain relevant issues. Not only would such briefings increase communication directly, but they
would also improve personal contact among different levels in the intelligence bureaucracies. Other
improvements in communication could be initiated informally by managers. These improvements
would require flexible managerial attitudes toward the organizational structure of the intelligence
community. For instance, mechanisms to improve contact between upper and lower level analysts
would require convincing upper level people of the importance of these contacts and lower level
people to participate freely without fear of negative career effects. Although these attitudinal
changes might prove difficult to institutionalize, they could be encouraged by building upon the
degree of informal exchange that already exists.
The most productive organizational changes in the intelligence community can be made within
the system, rather than by creating a new one. There are no absolute solutions to the problems
created by hierarchical bureaucracies, as a remedy to one problem will often create another. Hence,
we recommend that the intelligence community build on the frequent and important informal modes
of communication that already exist.
The conference also examined the impact of two other organizational issues: specialization and
centralization, and their impact on the analytical process. Specialization presents two separate
dilemmas for intelligence: First, though it is essential to the efficient command of detailed
knowledge required of analysts, specialization undermines the broad interpretative perspective
crucial to relevant analysis; Secondly, a high degree of specialization can block the sharing of
2
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
information, but has the important benefit of producing diverse opinions. If an analyst does not
reach an adequate level of understanding of his field, it is possible that important trends and
developments will not be detected and properly interpreted. An analyst who does not possess a
sufficient understanding of a particular situation may wrongly apply concepts and models derived
from other experiences.3 Or, his lack of an adequate background in, for instance, the history of a
region, might cause him to miss important insights that others with more substantial knowledge
would see. Thus a certain degree of specialization is crucial to quality analysis.
On the other hand, an analyst should not become so involved with the details of a particular
issue, that he loses sight of the big picture. Certain situations can only be assessed when viewed in
the context of a regional or global perspective. This broader perspective is especially important to
upper level analysts. Thus the dilemma: analysts must have a longtime exposure to one geographic
region or subject area in order to develop a sufficient background and level of expertise, but only
exposure to a variety of regions or subject areas produces the benefits of a broad perspective. The
intelligence community is currently trying to find the appropriate balance in asking such questions
as: How long should analysts remain in the same job? Should the agencies rotate them frequently
or leave them in the same position for an extended period of time?
2. We recommend that intelligence agencies offer analysts extended exposure to one region or
issue during the early stages of their careers and only then rotate their assignments.
Extended exposure to one country or issue is vital because we need true experts in a wide
variety of subject areas. Additionally, concentration on a specific area allows analysts to develop
knowledge and hone analytical skills, and can serve as a base from which to later develop a more
general perspective. These advantages are integral in training high level analysts and managers who
can look at analysis critically and confidently, especially as it relates to their particular areas of
concentration. Consequently, we feel that at lower levels, at least, analysts would benefit from an
extended exposure to one region or issue.
One way to avoid some of the problems associated with a high degree of specialization would be
to ensure recruitment of quality personnel and higher levels of training for analysts. Analysts who
3The issues of bias and mindset will be discussed at length later in the paper.
3
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
receive a thorough initial education, including basic theories of analysis, will prove less susceptible
to overly narrow perspectives than those trained on the job and they will have a better sense of what
is required of intelligence. It is also more likely that they will achieve a high level of expertise in a
particular subject more quickly and then will be able to rotate into related fields, while still
maintaining a high level of competence. This rotation after initial long term exposure to one field
will provide analysts with the opportunity to develop a broad view of the issues and alleviate some
problems caused by boredom.
One other potentially negative aspect of high degrees of specialization is that it may breed a
parochialism which often results in the formation of lines of secrecy and overzealous loyalty.
Consequently, rivalries may develop both between and within organizations based on the topical
focus of particular groups. Although one can argue that this type situation can provide a mechanism
for competitive analyses, and therefore better intelligence products, it might also result in a block in
the flow of data between groups. If specialization impeded the free flow of information to this
extent, it would result in weakened analysis. We find, however, that specialization is presently
responsible for very little parochialism within the intelligence community. Inter-unit rivalries may
well exist, but not to the extent that they mitigate against a proper level of analytical exposure to a
particular region or subject area.
The other important organizational pathology in the structure of an intelligence agency is the
necessary degree of centralization. Some centralization is needed to ensure proper coordination and
control of the intelligence bureaucracy. On the other hand, overcentralization may result in forced
consensus or cause too much intelligence information to overwhelm high intelligence officials with
data.4
Coordination of intelligence forces different analysts to question and compare views. This can
have good results if they reach a consensus through logical arguments, but runs the risk of
politicization of the intelligence product through the protection of parochial interests which supply
the policymaker with the "right" intelligence.5 It may also result in the suppression of legitimate
dissenting views. In the end, the degree of objectivity depends on the individual leaders in the
4Glenn Hastedt, "Organizational Foundations of Intelligence Failure," pp. 141-142.
5Michael Handel, "Strategic Surprise: Politics of Intelligence and Management of Uncertainty," in. Intelligence
Policy and Process, pp. 263-265.
4
h_ Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
intelligence and political communities and on the integrity of the intelligence. As a result, the
problems inherent in preparing group estimates may be impossible to address in any institutional
manner.
3. In order to reduce the problems associated with information overload, we recommend
filtering the flow of data up through the intelligence bureaucracies.
Minimizing the flow of data up through the system will best solve the problem of information
overload often associated with centralization. Although a filtering process may at times downgrade
the emphasis given to important pieces of information, the benefits far outweigh the risks. When
large quantities of information are sent through an organizational structure, top level analysts may be
overwhelmed by the sheer bulk of material that they receive and may not read it all with the proper
degrees of care and selectivity. As a result all information may well be lumped together and receive
inadequate attention. The events leading up to Pearl Harbor give the most publicized demonstration
of this phenomenon. Somehow, lower level analysts must pass on information to their superiors
under a system which assigns degrees of priority to different pieces of information or analysis,
whether through a formal grading system or some other less rigid mechanism. Such a system
would leave upper level analysts and managers to read anything at their discretion, but would offer
guidance as to what the lower level intelligence producers think is important. It would maintain a
high degree of flexibility for the upper levels in the bureaucracy, while helping to prevent the
negative aspects of data overload.
5
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Agency Specific Issues
The differences in the organizational contexts within which the various components of the
intelligence community operate provide insight into the problems of analysis that are specific to a
particular institution. According to Hastedt, the formal framework of institutional relationships "is a
goal-oriented system whose logic and coherence derive from the purposes or goals assigned to it by
forces outside the organization."6 The analytical products of each component of the intelligence
community differ, because they reflect to varying degrees assumptions and perceptions of the
national interest particular to each agency, the organizational process by which each operates, and
each agency's association with the policymaking community.7
The source of the major differences between the units in the intelligence community can be
found in the essential purpose of the organizations. The CIA, through its association with the DCI,
provides analyses geared toward the policymaldng needs of the President and the National Security
Council. INR is part of and provides analysis for the State Department.. These two organizations
operate with a strategic approach to analysis and their analysts are primarily concerned with the
political and diplomatic implications of state and regional issues. DIA, the intelligence element of
the Secretary of Defense and the Joint Chiefs of Staff, provides most of its intelligence to military
planners and operators. The DIA thus assumes a tactical approach to intelligence in order to answer
the questions of military capabilities, and the extent to which given operations affect the enemy.8
The individual units of the intelligence community define the same problem in different ways.
The way in which the policymaker perceives the problem therefore, affects the value that he attaches
to the various estimates. There was, for example, a division in the intelligence community over the
essential nature of the Vietnam War. The CIA based its analyses and estimates on the belief that the
war was a socio-political struggle, whereas the DIA saw it as a military problem. The DIA, as part
of the Defense Department, has the task of protecting the country against a military threat. In its
6Glenn Hastedt, "Organizational Foundations of Intelligence Failures," p.141.
7Stafford Thomas, "Intelligence Production and Consumption: A Framework for Analysis," in, Intelligence
Policy and Process, pp. 125-126.
8james Austin, "The Psychological Dimension of Intelligence Activities," in Intellignce Policy and Process,
pp.207-209.
6
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
search for military threats, it tends to overestimate the enemy's capabilities. The "threat inflation" of
the DIA works to increase military funding and thereby strengthen U.S. military capacity.9
Because the policymakers perceived the war in essentially military terms, the military analyses of the
DIA had greater influence than the more politically and diplomatically slanted views of the other
agencies.
In contrast to accusations of DIA's "threat inflation," INR is often said to underestimate the
consequences of a particular situation. This is primarily a result of the diplomatic nature of the State
Department. Many INR analysts are foreign service officers who previously served in the countries
in question, and often aspire to become ambassadors. It is natural for such people to rely heavily on
diplomatic solutions to problems. Concluding that there is no diplomatic solution, in effect
questions the capabilities of the foreign country's leaders. By reducing the credibility of the people
he used to associate with, a foreign service officer somehow reduces his own credibility as well.10
Therefore, a combination of the diplomatic nature of State Department activities and the past
experiences of its personnel has resulted in generally forgiving estimations of enemy capabilities and
intentions.11
1. While it is important to educate policymakers about possible bias in the work of the various
intelligence agencies, no attempt should be made to correct the institutional biases of the three main
analytical agencies.
Attempts to "correct" the institutional biases inherent in the intelligence community will
ultimately prove fruitless and counter-productive. These organizations have evolved so as to
maximize their abilities to fulfill their respective roles in the community. There should not be a
community-wide attempt to combine all of the positive elements of the different intelligence
organizations. Each institution has developed certain traits which enable it to fulfill its role in the
community. Removing the variability between the organizations would eliminate the existence of
alternative approaches to intelligence. Institutional biases are relevant to the various agencies' roles
9S tafford Thomas, "Intelligence Production and Consumption: A Framework of Analysis, " p.134.
10This tendency to maintain overly optimistic views about the status of a country in which one was formerly
stationed is known as clientism.
11From an interview with a former CIA National Intelligence Estimates Officer.
7
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
in the policymaking process and thus should be preserved. Having examined the beneficial effects
of institutional biases on the analytical process, we now turn to an assessment of the specific
characteristics peculiar to the three intelligence agencies covered by the Conference.
The Central Intelligence Agency
There is little in the organizational ethos of the CIA that requires a specific tailoring of the
general observations presented thus far in the report. We discuss one unique feature of the CIA,
covert operations, in a separate section. The other element that distinguishes the CIA from the rest
of the intelligence community, is that its director also oversees the community as a whole. This dual
function gives him a crucial role in determining the quality of various aspects of intelligence.
The Director of Central Intelligence (DCI)is the most powerful person in the intelligence
community. He functions as the primary conduit in the transfer of the intelligence product to upper
level policymakers and is responsible for tasking the intelligence community. Additionally, he
serves a crucial role in the intelligence budget formation process. The intelligence community, and
particularly the CIA, often reflect and at least must react to the personal preferences of the DCI.
Consistent with the historical personal influence of the DCI, the community continues to maintain a
degree of flexibility that permits it to respond to its director.
The National Security Act of 1947 and the Central Intelligence Act of 1949 established that a
Director of Central Intelligence would head both the CIA and the intelligence community as a whole.
These dual roles of the DCI have caused continual controversy. Some argue that we should divide
the responsibilities of the DCI between two separate offices, a director of the CIA and a Director of
National Intelligence (DNI), in order to reduce the chance of bias in intelligence reporting and clarify
the lines of authority and responsibility within the community. While this division of
responsibilities might help solve certain problems, the current system is on the whole, superior.
8
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
2. We recommend preserving the dual roles of the DCI.
"Since its creation, the CIA has provided a power base for the director of Central Intelligence as
head of the agency, sustaining him in the separate effort over the years of exercising more and more
coordinating authority over the intelligence community." 12 This bureaucratic power base within
the CIA allows the DCI to function far more effectively than he would had he no organization of
support on which to rely. The DCI's ability to communicate with top policymakers and his
influence in coordinating the various components of the intelligence community depends directly
upon the authority invested in him. Thus, removing his organizational support could well
emasculate him bureaucratically.
Creating an office of the DM would also necessitate the establishment of a new, largely
redundant bureaucratic layer. Further, as the CIA generally provides the analytic resources for the
DCI to fulfill his obligations as chief intelligence officer to the President, removing the DNI from
this source of information would undercut his influence with policymakers and reduce his capacity
for relaying analysis. Thus, while dividing the roles of the DCI may seem appealing, the overall
effect would be to reduce the power and capabilities of the DNI to such an extent as to render him
significantly less effective. The DCI presently holds a position of tremendous influence within the
foreign policy community. In order to function effectively, he must have a level of bureaucratic
support and organizational power commensurate with his responsibilities.
3. The proper degree of authority is presently invested in the DCI.
In order to be effective, the DCI must have sufficient authority to coordinate the intelligence
community budget, administration, collection requirements, and national intelligence estimates
(NLE's). Until the Nixon administration, the DCI was often hampered by lack of explicit authority,
12John Ranelaugh, The Agency: The Rise and Decline of the CIA. (London: Weidenfield and Nicleson, 1986),
p. 676
9
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
4
although DCI's sometimes overcame this lack of official authority through personal expansion of
their jurisdiction.. Executive Orders issued since have addressed this issue to a large extent. The
Reagan years have witnessed a slight decrease in the DCI's degree of authority over the community,
but, in general, the DCI's level of authority gives him adequate ability to monitor the collection of
intelligence information, task the various agencies in the community, determine present and future
priorities, and coordinate the budget.
4. The DCI must function as an unbiased reporter of intelligence analysis, not as a policy
advocate.
In his most important capacity, the DCI facilitates producer/consumer interface at the highest
levels. As the nation's chief intelligence officer, he bears primary responsibility for coordinating the
NIE's, providing the President with intelligence analysis, and advising the National Security
Council. The DCI coordinates the tasking and formulation of the NIE's through the National
Foreign Intelligence Board (NF1B), an organization composed of the National Intelligence Officers
(NI0s). The NLE's pool a variety of resources from the various intelligence agencies. "It is the
task of the Director of Central Intelligence, utilizing his influence in the various interdepartmental
mechanisms, to create from these diverse components a truly national estimate, useful to the national
interest and not just to a particular bureaucratic interest." 13 The DCI must not, as former Director
William Casey was reputed to have done, slant the reporting in the NIE's to favor a particular
policy. Rather, he must ensure that the NIE's remain unbiased, analytical reporting. Preserving the
objectivity of the NIE's is further magnified by the fact that it is the DCI who presents the NIE's to
the NSC.
Though a statutory advisor to the NSC, the DCI has become a de facto member of the Council.
The DCI not only presents the NIE's to the NSC, but also briefs the Council at regular intervals
13 Ransom, Harry H., The Intelligence Establishment, (Cambridge Mass: Harvard University Press, 1970),
p.276
10
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
with intelligence synopses. Additionally, the DCI serves on two NSC subcommittees- the Senior
Intelligence Group for Intelligence (SIG-I) and the Special Situations Group (SSG). This active
role on NSC subcommittees provides the DCI with another, more informal means to insert
intelligence analysis into the policymaking process. These important ties heighten the danger of
placing the DCI too close to policy, thus biasing his reporting.
5. The post of DCI should not be elevated to cabinet level.
The DCI already serves as the President's chief intelligence advisor. As such, he must walk a
'fine line between ready access to the President and over active involvement in the policymaking
process. In order to function properly as the President's advisor, the DCI must have a degree of
flexibility to adopt his own, personal relationship with the President. One restriction, however,
should become law. The President must not appoint the DCI to his Cabinet, as this appointment
jeopardizes his effectiveness as an unbiased reporter of intelligence by institutionalizing his role in
policy formation. The DCI must retain his perspective as an advisor to the policy making process,
not an as advocate or supporter of policy.
6. The DCI must be allowed adequate flexibility to shape the intelligence bureaucracy as he sees
fit.
The character, personality, and background of the DCI directly affects his ability to both lead the
CIA and the intelligence community and to transmit intelligence to policymakers. The system has
remained flexible enough to allow each DCI to shape it according to his own predilections and to
stress the issues that he deems important. Maintaining this flexibility in the system contributes
substantially to the effectiveness of the community. Each DCI must continue to have the ability to
11
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
4.
tailor it to his beliefs. This is not to say that there is no bureaucratic resistance to change within the
intelligence community. The experience of Stansfield Turner certainly demonstrates that. The DCI,
however, can affect the bureaucracy to a certain extent and we should not set up a rigid formulation
that inhibits this flexibility.
7. Historical observations of past DCIs should serve as guides to future appointments.
A historical perspective allows us to isolate several important traits in a DCI. Most important,
the DCI "should be a man with a political base outside the Agency." 14 Without some sort of
power base outside the Agency, it may prove difficult for the DCI to establish a perception of his
importance outside the intelligence community and to gain adequate access to high level intelligence
consumers. Such a base also has traditionally contributed to the DCI's effective coordination of the
various branches of the community. Additionally, political influence independent of that derived
from the administration or the intelligence community will help the DCI to resist political pressure
from intelligence consumers to politicize the intelligence product. Those Directors who have risen
through the ranks of the intelligence service have functioned especially well as managers of the CIA,
but have generally lacked the influence needed to effectively manage the whole intelligence
community and have had difficulty gaining real access to the President.
In 1984, Senators Moynihan and Goldwater proposed S 3019, stating that the DCI and the
DDCI "should be appointed from among career civilian or military intelligence officials." 15 Such
statutes, however, do little to ensure that the best individuals will fill those positions, while limiting
the President's flexibility in appointments. We should look to history for general guidelines,
without rigidly specifying the credentials of a DCI.
The Rockefeller Commission Report concluded that "experience in intelligence service is not
14Powers, Thomas, The Man Who K-,.pt the Secrets,(New York: Alfred A. Knopf, 1979)*, p.8
15Annual Report of the Senate Select Committee on Intelligence, 1984, P. 28
12
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
?
necessarily a prerequisite for the position; management and administrative skills are at least as
important as the technical expertise which can always be found in an able deputy." 16 The
Commission further stressed the importance of personal characteristics such as integrity, stature,
courage, and independence. These attributes, while certainly important, do not encompass all the
abilities and qualities that a DCI must possess. He should also have a strong understanding of
foreign policy and, if possible, some background in foreign affairs to allow him to understand
policymakers needs and perspectives. Additionally, he must have a strong, complementary
relationship with his DDCI. Appointment of a DDCI with strengths different from those of the DCI
produces an effective team at the top of the community. Today, Judge Webster as DCI and Robert
Gates as DDCI, who bring different backgrounds and strengths to the relationship, should form an
effective managerial team.
From a general perspective, it is essential to maintain the flexibility of the DCI to perform his
tasks as he sees fit. That said, certain broad propositions should serve as guidelines for the DCI's
role. He should have a strong power base both within the CIA and preferably outside the
intelligence community. He must have easy access to the President and top policymakers, yet he
must guard against becoming a policy advocate. Finally, history can offer us certain characteristics
that have served DCI's well in the past. We should look for similar attributes in future DCIs.
The Defense Intelligence Agency
We propose a rather drastic restructuring of the Defense Intelligence Agency. As a result, we
discuss the peculiar features of the DIA at length in this section in order to make clear the reasoning
behind our departure from the theme of fine tuning rather than structural change that dominates the
rest of our report.
We can trace the fundamental problems in the DIA to the circumstances of its formation and
16Commission on CIA Activities Within the United States,Rockefeller Commission Report, 1975 p.17
13
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14 : CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
development. DIA's history divides cleanly into four phases: the consolidation of management
authority, a crisis of credibility, its response to critics, and its maturation. Established by a
Department of Defense (DOD) Directive, the Agency lacked the powerful mandate that accompanied
NSA's formation by Presidential decree. When he created the DIA, Secretary McNamara purposely
furthered the popular centralization trend of the late fifties, characterized by the 1958 reorganization
of DOD. He hoped to introduce the new agency without disrupting the important ongoing functions
of the military intelligence community.
OSD issued three guidelines to the first DIA director. The outfit (1) had to improve the
responsiveness of defense intelligence to decision makers and increase the efficiency with which the
community employed its resources; (2) assume new responsibilities on a specific, calibrated
schedule; and (3) leave certain functional responsibilities with the services until a later date, when
their transfer would be carefully reviewed. This incremental implementation denied DIA the
bureaucratic momentum that might otherwise have accompanied the formation of the new agency.
The Director of DIA (DDIA) thus began operations with a sense of caution, rather than with an
aggressive mandate.
DIA's position in the intelligence community posed two serious problems. First, the JCS, one
of the Agency's two masters, opposed its formation. The Secretary of Defense had rejected the
JCS 's plan for a Military Intelligence Agency that would report only to the service chiefs. DIA
arose from a compromise that diluted its influence with the SECDEF by making it responsible to
him through the JCS; nevertheless, the Chiefs opposed DIA because it transferred activities from
their personal jurisdictions to collective control, thereby dissipating their power. There was also a
risk that moving service personnel to DIA would decrease the responsiveness of intelligence to each
service's tactical needs. Thus, the services jealously guarded their top performers and combat
leaders, preferring to send individuals from support branches to DIA. In fact, the common
complaint became the lack of quality people at DIA, not in the independent military branches. This
perception tainted military employees throughout DIA's first decade.
Second, DIA had existing rivals in the field of strategic military intelligence, primarily CIA,
though to a lesser extent NSA and INR as well. Competition between pre-existing analytical
14
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
4.?
bodies can be productive, since it may inspire both to higher standards. Unfortunately, when a
new agency must compete against entrenched rivals in hiring personnel, it may never build the solid
core of people that it needs to establish its credentials with policy makers. Arguably, the latter
situation occurred in DIA.
Because of these conditions, efforts to consolidate internal and external management authority
characterized the first phase of DIA's development, which lasted until 1970. Within that time DIA
underwent four reorganizations. It faced an uphill battle to centralize defense intelligence, because
after Congress refused to construct a headquarters for it, DIA operated from buildings spread across
the Washington area. By 1965, OSD had transferred to DIA all of the functions that the Office had
considered moving there in 1961. These activities included intelligence training, scientific and
technical intelligence, MC & G (mapping, charting, and geodesy), and the Defense Attache System
(DAS). It was during this time that that DIA established the Defense Intelligence College (DIC) to
train attaches and defense analysts. Although the reassignment of these duties reflected growing
confidence in DIA during its period of consolidation, new obstacles arose to delay its maturation.
The 1970 Blue Ribbon Panel on Defense, the Fitzhugh Panel, concluded that DIA had- "too
many jobs and too many masters." 17 The SECDEF and JCS often had jurisdictional disputes.
Since DDIA reported to both of them as an advisor and staff officer, his divided loyalties sometimes
caused delays or guidance that was less than candid. The Panel offered no concrete reforms on this
matter. DDIA was also caught in a command loop. Although responsible for integrating the service
intelligence branches, DDIA "reports directly to the Joint Chiefs of Staff, comprised in four-fifths
majority by the Senior Officers of the four Military Services. ..whose intelligence programs [he is
charged with coordinating].
18
Despite OSD's intent to centralize defense intelligence under DIA, the services rebounded with
astonishing speed. At the time of the investigation, the services' agencies had larger staffs than
before the transfer of responsibilities to DIA. As the Panel observed: "It is paradoxical that DIA
cannot develop a capability to perform its assigned functions while the Military Departments, which
provide a large proportion of DIA personnel, maintain the capability to produce intelligence
estimates." 19
17U.S. Blue Ribbon Panel on Defense, Report to The President and the Secretary of Defense on the Department
of Defense (Washington: GPO, July 1970), p.45.
I8Ibid.
15
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
The Fitzhugh Report, along with the Schlesinger Report in the following year, resulted in the
formation of the Defense Mapping Agency and Office of Net Assessment. The former agency
relieved DIA of a labor-intensive task, whereas the latter became a new consumer of DIA's
products, adding to the Agency's demands.
As the US withdrew from a lost war in Southeast Asia, public opinion hardened against the
intelligence community and the military. Not surprisingly, the DIA came under heavy fire during
the early 1970s. These years marked the second phase--a period of turmoil--in the Agency's
development. Nowhere were DIA's shortcomings more apparent than in the estimates process.
Critics accused DIA of falsifying threat estimates in Vietnam and inflating Soviet assessments to
bolster DOD budget requests. 20 To add to its difficulties, DIA has always had to produce more
detailed military analyses than its non-DOD counterparts' because its military customers had
operational responsibilities.
Internal problems also plagued DIA. According to the then Assistant Secretary of Defense for
Intelligence, analysts were still dispersed across the District with "unsatisfactory communications
linkages." 21 The personnel system had stagnated with mediocre personnel filling many posts. As
of 1971, not one DIA analyst had achieved a supergrade rating and no military officer assigned to
DIA had been promoted to flag rank.
Although he could do nothing about inequities in promotion, General Donald Bennett, DDIA,
directly confronted the deficiency in estimates by establishing the Directorate of Estimates (DE) in
1970. DE concentrated a number of analysts in a relatively isolated office in Rosalyn, Virginia,
where they had the freedom to concentrate on long range issues. In his capacity as Deputy Director
for Estimates, LTG (then MG) Daniel 0. Graham began a campaign in the open literature to
establish DIA's credibility in the field of estimates. His key article, "Estimating the Threat: A
Soldier's Job," appeared in a 1973 Army magazine. Graham's central point was that policy makers
inside and outside of DOD regarded all military assessments as self-serving and consequently
19Ibid.
20Patrick J. McGarvey, "DIA: Intelligence to Please," in Morton H. Halperin and Arnold Kanter eds., Readings
in American Foreign Policy, (Boston: Little, Brown, and Company, 1973), p. 324.
21Dr. Albert C. Hall, quoted in "Better Deal for Spooks?," Armed Forces Journal International 112, (April
1975): p. 12.
16
L_ Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
deferred to civilian sources. After describing DE's contributions, he concluded that:
The military profession should reassert its traditional role in the function of
describing military threats to national security.. ..The best assistance the
Army can give to such an effort is to insist on objective strategic
intelligence, cooperate with DIA in producing it, and put good officers in
the strategic intelligence field. 22
His observation was accurate but, as further evidence indicates, largely unheeded by the
services.
DIA's period of turmoil culminated in the 1976 congressional inquiries, in which both the
Senate and House investigative committees contemplated DIA's abolition. While the Senate's
Church Committee concluded that the Agency's elimination would be counterproductive, the Pike
Committee recommended it. The latter group opted to publish only its recommendations, not the
text of its hearings and internal debates, prior to the executive branch's sanitizing them, a decision
which sparked much controversy.
The Village Voice published an unauthorized version of the Pike report. According to that
version, the House committee decided that there were two problems inherent in DIA's
civilian/military structure that required drastic action. First, as long as the services retained their
respective intelligence branches, DIA would remain an unattractive assignment to promising military
officers. Second, DIA was unable to offer civilians opportunities for advancement and
responsibility that were competitive with CIA and NSA because of the military's monopoly on
upper management positions.
In the mid-seventies, DIA entered its third stage of development. The Agency fought back
against critics and profited from the reforms that were instituted during the lean years of the early
1970s. Upon his retirement as DIA director, LTG Graham publicly criticized the House
committee's proposal to kill DIA, calling it a "strange recommendation. Neither I nor any of my
deputies testified before the (Pike) committee on DIA. The only time I testified was on the Tet
business, the VC/NVA estimates." 23 He claimed that junior personnel from the staff were "turned
loose" for six weeks within the organization yet failed ever to grasp that it was an analytical facility,
not a clandestine collection body.
22MG Daniel 0. Graham, "Estmating the Threat: A Soldier's Job,: Army 23, (April 1973): p. 18.
23F. Clifton Berry, "Former DIA Director Speaks Out," Armed Forces Journal International 113, (March 1976):
p. 15.
17
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
In 1976 a civilian writer, Hanson Baldwin, reinforced Graham's defense of DIA by arguing that
the Central Intelligence Agency had become too large and acquired too much power, often assuming
an adversarial relationship with the DIA or other agencies. To avoid an "intelligence party line" on
major issues, Baldwin argued that DIA should be strengthened relative to the civilian agencies.
The Agency did grow stronger, but not from external support. Instead, a partial response to the
Pike report emerged de facto, as tours of duty at DIA became an accepted part of the career path of
all military intelligence personnel. According to LTC Charles Lovejoy, a recent instructor at the
Defense Intelligence College (DIC), most Army intelligence officers view a stint at DIA or its
training facility, DIC, as a normal step in career progression. 24 A CIA official who served in DIA
during its infancy disagrees, but is willing to concede that an individual's perceptions and the actual
effect of such an assignment vary according to one's specific job at DIA. 25
According to DIA Historian Deane Allen, the fourth stage of the Agency's existence began in the
early eighties, when the organization reached institutional maturity. The last two DDIAs have been
the first leaders who could concentrate on setting defense intelligence policy without having to be
protective of the Agency. In FY 1981 and 1982, Congress filially appropriated funds with which to
construct a headquarters facility for the Agency. The physical proximity of departments in the new
building has improved morale and efficiency.
Despite improvements in its credibility and performance, the Defense Intelligence Agency
operates far below its potential. First, there is a faulty chain of command in the defense intelligence
system. DIA still suffers from "too many masters," as the Fitzhugh Panel described in 1970. The
Director of Central Intelligence, the Secretary of Defense (SECDEF), the Joint Chiefs of Staff
(JCS), and the service intelligence branches, all exert influence on the Agency. While these bosses
all make demands on DIA, they do not necessarily utilize its products. The SECDEF, for instance,
has traditionally relied heavily on CIA, using it to check DIA's reports. The Director of DIA's
(DDIA) three star rank contributes to the problem, as he has relatively little institutional clout. When
he needs to pressure one of the service intelligence chiefs, the DDIA must complain to that service's
24LTC Charles D. Lovejoy, Jr. in an interview at Princeton University on October 27, 1987.
25Interview with a CIA official at Princeton University on October 15, 1987.
18
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Chief of Staff, a four star officer. In practice the DIA Director derives his influence largely from
personal relationships with commanders of the service agencies. The problem is that future DDIAs
may not enjoy close ties.
Second, DIA controls very few collection resources relative to the requirements that it bears.
DAS is the only such program that DIA directs. Despite some excellent attaches, the system does
not produce high quality, consistent intelligence commensurate with its costs. An INR official
complained that while recent data from the Philippines has been extremely insightful, reports from
the Republic of Korea failed to address the important question of how the Korean military would
react to an opposition victory in elections.
Third, the quality of DIA's personnel has never rivalled that of CIA, NSA, or INR. Once
institutionalized, the personnel problem has perpetuated itself in the same way that athletic
recruitment remains poor for schools whose teams have losing records. The slow promotion rate
within the military for intelligence officers has inevitably convinced many top performers to choose
a combat arms branch. Furthermore, the services cooperate with DIA in widely varying degrees.
Although the original plan was to rotate the directorship between the services, there has been only
one complete circuit. DDIA has alternated between Army and Air Force officers for the last six
tours. The Agency has essentially become an Army and Air Force fiefdom, with the Navy sending
those people whose absence will least affect its operational needs. Of the military personnel at DIA
in 1985, 39% were Army, 36% Air Force, 23% Navy, and 2% Marine Corps. According to one
DIA analyst, Navy personnel are usually Lieutenant Commanders or below, females, or officers
destined never again to hold an intelligence assignment. While only the last of these attributes is
negative, the conglomeration reflects the Navy's unwillingness to send experienced, unrestricted-
line officers to DIA.
Problems with civilian personnel have different origins. Although in 1962 Congress approved
an Excepted Service Appointment System, it was a long time before DIA received Civil Service
(CS) exemptions equivalent to CIA's. Until 1985, a popularly suggested reform for DIA was to
19
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
rectify this imbalance. There was a consensus that improvement in civilian analysts could come
only when DDIA had broader powers to promote and dismiss the civilians. Apparently, there were
two rationales for exemption. First, without CS restrictions on salaries and positions, DIA could
fashion a management system more responsive to its needs. It could establish more mid-level
management slots for civilians, thereby increasing their incentive to remain with the organization.
DDIA could dismiss or demote unproductive senior employees to increase advancement
opportunities for new people. Second, security considerations have dictated that the firing of
intelligence personnel should not be appealable to external sources like a Civil Service review board.
The Carter Administration broadly addressed the first problem by creating the Senior Executive
Service, which increased the number of supergrade positions in government in order to retain
experienced civil servants; however, Congress did not allow DIA to participate in the program until
the FY 1985 Authorization for Intelligence Activities, in which it established the Senior Defense
Intelligence Executive Service. It was recently decided that DIA's Executive Director, who
"exercise[s] day-to-day management of internal Agency operations and activities," should
henceforth be a civilian. 26 Although it merely institutionalizes the status quo, this decision
guarantees opportunities for advancement to aspiring civilian managers, while assuring continuity in
the third highest Agency job. It is too early to assess the effect of these measures on civilian quality
problems.
Analysts both inside and outside the Agency contend that while it performs well in certain areas,
DIA lacks the experienced analysts found in CIA and INR. It is not uncommon for young analysts
to join DIA at the beginning of their careers, gain some experience, and then transfer to Langley
where their reputations and responsibilities can expand more rapidly. An analyst observed that
shortages of trained specialists force the Agency to assemble haphazard collections of people to meet
unexpected intelligence requirements. When the Falklands crisis erupted, DIA drew people from
places as diverse as its Soviet and East Asian desks, just to meet its manpower needs. Given such
conditions, it is not surprising that analysts depart for other agencies with less turbulent
26U, Department of Defense, Defense Intelligence Agency, DIA: Organization. Mission. and Key Personnel,
(Washington: GPO, 1986), p. 6.
20
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
environments and where their experience can more regularly accumulate.
For the most part, the problems described above stem not from DIA, but from its disadvantaged
position relative to other intelligence organs. To maximize the effectiveness of both human and
material resources, the Department of Defense must not allow DIA to stagnate. Decision makers
should adopt the following changes in order to ensure the Agency's continued viability.
7. The Military Intelligence Community should be significantly restructured.
The conference recommends that the Defense Intelligence Agency be made the military
intelligence center for the American government by abolishing the present service intelligence
branches in favor of a single, unified Military Intelligence Service (MIS) that reports only to the
SECDEF. The DDIA, elevated to a four star position, will command the new group, as well as the
Agency. The three star Deputy Director for JCS Support should assume full responsibility for the
demands of the JCS.
Consolidating the service branches within the confines of DIA will vastly simplify coordination
and oversight. The elevation of DDIA from 0-9 (LTG) to 0-10 (General) with joint command of
DIA and MIS will increase the Agency's actual and perceived authority. The transfer of the
services' personnel and resources also increases control over collection, while ensuring that the
armed forces cannot easily reconstitute tactical intelligence branches. The services, and therefore the
JCS, will have to depend heavily on DIA for the first time. DDIA can then concentrate on servicing
SECDEF without succumbing to bureaucratic pressure from the Chiefs, who will no longer outrank
him.
Opposition to the proposal will come primarily from the services. With this barrier in mind, the
first step towards a revitalized DIA is for the President to commission a detailed study on the
implementation of this proposal. The lesson learned from the Agency's formation is that major
defense reorganization requires an executive mandate. As both Chief Executive and Commander-
21
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
in-Chief, the President wields power over the JCS and their intelligence branches that Office of the
Secretary of Defense could never muster. As long as independent intelligence units exist in the
services, the services will not cooperate fully with DIA, to the detriment of policy makers. ADM
Stansfield Turner recognized this problem in his book:
As long as there are four separate service intelligence organizations that
command higher loyalty and are not so bound up in bureaucracy, the
better analysts are going to avoid going to the DIA. In my view, only if
the service intelligence organizations are disbanded permanently is there
hope for measurably improving the DIA's capabilities. There is no
serious need for those four separate organizations that the DIA could not
fill. 27
8. DIA should focus on improving the quality of its personnel.
DIA can increase the level of experience of military intelligence personnel by enacting the
following changes. MIOs (Military Intelligence Officers, members of MIS), will serve in one of the
four services until they reach 0-3 (Captain in Army, Air Force and Marines, and Lieutenant in
Navy), at which point they may apply to join the unified military intelligence organization. If
accepted, they will transfer their commissions to MIS, and henceforth belong to an independent
service, complete with a new uniform. MIS will determine their promotions, assignments, and
salaries.
To avoid paralyzing the defense system, MIS should leave its new employees, including 0-1s,
0-2s, and enlisted personnel, at their current duty stations, but issue new uniforms and quickly
assert the new chain of command. The new MIOs will maintain an affiliation with their former
service by virtue of their early military experiences, as well as their frequent duty assignments there.
These factors should insure a healthy commonality of interest with the forces that the MIOs support,
but their lack of career ties should give them a more objective view of their mission. The job of an
27Turner, Stansfield, Secrecy and Democracy, (Boston: Houghton Mifflin Company, 1985), p.247
22
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
MIO should be to provide the best possible military intelligence to decision makers in both
government and the operational military branches.
9. The Military Intelligence Service should implement new programs for continuing education.
To attract and retain high quality military and civilian employees, a large-scale program of DIA
fellowships should be instituted. Available to both civilians and MIOs, these scholarships should
encompass graduate studies, exchanges with other intelligence agencies, or mid-level posts
elsewhere in the national security apparatus. MIOs should spend at least every fourth year on
assignment outside of DIA. Of course, to fill standard intelligence slots in combat commands,
many MIOs will be operating with their parent services at any given time.
New opportunities for continuing education should improve the accession and retention rates for
both civilian and military employees. In the long term, investments in education pay higher
dividends than salary increases, because they simultaneously increase human capital and the job
satisfaction of ambitious employees. The resources saved by better interservice coordination could
finance the program of organizational fellowships.
The Bureau of Intelligence and Research
The bureaucratic ethos of INR is characterized by close working relationships with the policy
side of the State Department. As a result of the diffusion of power in the State Department, INR has
a special responsibility to disseminate intelligence at all levels. In addition, organizational factors,
personnel composition, and intelligence audiences, specific to INR, contribute to INRs relatively
high degree of influence with the its consumers.
INR is divided into regional and functional offices which parallel the regional and functional
bureaus on the operational side of the State Department. This structure maximizes INR's contact
with policymakers. The staff of each office at INR interacts daily, both formally and informally
with its counterpart, the operational bureau. The core of INR is the country analyst. On a daily
23
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
basis, this analyst monitors incoming data, "flagging" anything important, and makes daily reports,
which his immediate superior reviews. The position of country analyst parallels that of desk officer
in the regional bureaus of the operational side. While the analyst reviews intelligence and produces
analysis for policy-makers, the country officer deals with the administrative needs of the embassy,
and the details of policy implementation. These officers work together both formally and
informally.
10. INR should preserve and enhance close contact between desk officers and country
analysts.
Close relationships between desk officers and the country analysts have both benefits and
liabilities. The desk officer is of unique value to the analyst in that he is the Embassy's primary
contact in Washington. The desk officer handles all the embassy's official and unofficial
communications with Washington and is often aware of policy decisions and disputes within the
Department. In order to be maximally effective the analyst needs these informal contacts in the
operational bureaus. INR's access to policymakers even proves to be an informal resource for the
more isolated CIA analysts. According to INR analysts, CIA analysts often ask them for clues to
policymakers' needs.
Just as the analyst relies on the desk officer for policy and embassy gossip, the analyst can feed
the desk officer intelligence. If there is a piece of analysis that directly affects a desk officer, the
analyst usually will take a "bootleg" copy to the bureau the day the item comes out rather than
waiting for a weekly meeting. In the relationship with the desk officer, the country analyst has an
opportunity to "sell" his intelligence to an eager consumer. These forms of close, informal contact
between analysts and policymakers contribute to the timeliness, relevance, and influence of INR's
work.
The primary liability of the close relationship between the country analyst and the desk officer
24
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
particular to the Foreign Service is known as "clientism". INR analysts may demonstrate pervasive
biases toward the "client" state. A certain degree of bias need not damage their work. As members
of the State Department, working constantly in this atmosphere which emphasizes the "client state",
INR analysts should adopt the attitudinal ethos of those with whom they work. In particular, the
pervasive understanding of cultural and historical backgrounds makes INR analysis distinct.
Assuming that the State Department's analysis will compete with other viewpoints in the IC, there is
merit in such a bias. On the other hand, INR must balance cultural understanding or personal
knowledge of foreign officials, and "clientism". The Foreign Service Officer turned analyst in
particular must beware of these biases in his own analysis.
The proximity of the country analyst to the desk officer allows him to evaluate "clientism" in
operational officers and combat it by providing intelligence which will broaden their understanding
of a situation. The influence of the country analyst on the desk officer as a second opinion and a
counterbalance to the reports of the embassy merits the continuation of this close relationship.
Analysts at INR say that a successful relationship between the desk officer and the country analyst
depends on "agreeing to disagree" Formal and informal contact with the rest of the intelligence
community helps to reinforce the perspective which INR analysts must maintain relative to
policymakers.
11. INR should continue to include Foreign Service Officers among its analysts.
As a result of reforms undertaken in 1954-55, the staff of INR was integrated to be 50% civil
service and 50% foreign service. Foreign Service Officers have 2-3 year assignments to INR as
analysts, and then return to the bureaus or to overseas assignments. Civil servants in INR who are
career intelligence officers have no obligations to the regional bureaus and such analysts tend to
remain fairly independent. Intelligence training and experience, and the constant exposure to the
25
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
rest of the intelligence community balance these civil servants' close relations with operational and
policy issues.
The problem of "seepage" arises out of the career patterns of the foreign service. FSOs serving
as analysts return permanently to the operational bureau whose policy they must criticize during
their days at INR. Such analysts have strong to incentive soften their analysis. The integration of
foreign service officers, therefore, may jeopardize INR's neutral and provocative analysis. In
addition to the already strong tendency to defer to the operational bureau, incoming FSO's may
themselves have a bias toward "clientism", and are certainly less likely to recognize it in others.
Nevertheless, removal of FSOs from INR would have several disadvantages. First, FSOs
would not receive the intelligence training which they badly need. Intelligence experience gives
FSOs a perspective which diminishes the effects of "clientism" on the policy side of the State
Department. Second, the perspective of FSOs as producers and implementers of policy positively
complements the perspective of the Civil Servants who do only intelligence work. Third, INR, in
order to maintain its influence with policymakers must be an integrated part of the Department. To
make INR solely civil service might isolate it from the informal policy processes in the regional
bureaus. Thus while the presence of FSOs in INR does pose certain threats to analysis in the
bureau, close contact with the policymaking community and concomitant influence within the State
Department outweighs these concerns and we suggest that the current structure be retained.
12. INR should place more emphasis on pro-active and long-term analysis.
As a result of its close proximity to policy makers, INR is uniquely suited among the
intelligence agencies to pro-active and long-term analysis. Intimate knowledge of the policy process
provides INR analysts with the opportunity to raise questions pertinent to the policy process.
Self-generated and long-term work creates perspectives which policymakers might not otherwise
consider. This is not to say that INR should become a policy advocate. The neutrality of the bureau
26
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14 : CIA-RDP90G01353R001700010025-0
is essential to its credibility within the Department. It is, however, important to recognize that
INR's proximity to policymakers affords it the special ability to encourage their thoughtful
consideration of issues. Intimate knowledge of the policy process allows INR analysts to identify
flaws in the thinking processes of the policymaker and insider status allows INR analysts to address
these flaws through both formal and informal mechanisms.
INR, however, does not presently have the resources or the personnel to fully exploit this
capacity. INR must focus its work on current reporting in order to satisfy and maintain its influence
with its State Department audience. This limits INR's ability to do pro-active and long-term work.
Analysts at INR have neither the time nor the resources to pursue the insights which proximity to
policymakers affords them. At present, self-started projects receive a low priority. In order to fully
exploit its proximity to policy and policymakers concerns, INR should receive increased resources
earmarked to exploring self-generated projects.
Emphasis on pro-active and long-term intelligence will inevitably create some tension with all
levels of policymakers in the State Department. INR's direct access to the Secretary should ensure
that the operational bureaus cannot easily override INR's "second opinion". Nevertheless, INR
must also maintain its good reputation as an intelligence agency rather than a policy critic in order to
ensure a receptive audience in the operational bureaus and thus "pro-active" work must remain
secondary to the current analysis in which INR excels. INR should exploit its capacity for pro-
active and long-term work carefully in order to balance their benefits against the potential costs of
politicization and possible loss of influence.
13. INR should be classified as a National Security Agency.
Classification as a National Security Agency would ensure INR a higher priority among federal
27
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
agencies and would separate INR from the State Department in the appropriations process. Such
approval would recognize the Bureau's excellent service, increase its resources, and guard it
against spending cuts at the State Department. In order to maintain INR's primary obligation to the
Secretary of State, however, the DCI should recommend appropriations subject to the Secretary's
approval. Although changing INR's budget in this manner might somewhat endanger INR's status
as an in-house aiency by isolating it as a bureau, the presence of Foreign Service Officers in INR
and the agency's physical position within the Department should eliminate this possibility.
Problems Inherent in the Nature of Analysis
In addition to organizational and bureaucratic impediments to high quality analytical work, there
exist certain factors in the nature of analysis itself that present potential problems. Intelligence
analysis is a human process and thus prone to human error. Although most of the time analysts do
produce high quality work, occasionally the human susceptibility to error results in mistakes. That
is not to say that certain errors cannot be avoided, but merely to recognize that no system can
eliminate all mistakes. In this next section we examine the analytical process to isolate potential
problems and recommend solutions where possible.
Consensus-Building versus Competitive Analysis
The issue of consensus-building arises whenever a group of people with varying opinions have
to produce one joint product. In this situation, compromise can become more important than the
search for "truth," and intelligence estimates produced in a consensual manner can become victims
of "group-think."28 The danger in consensus-building lies in this perversion of the original
purpose of joint research, which is to bring together experts in different areas to analyze all aspects
of a situation in arriving at a conclusion. What may happen is that participants with similar views
may form coalitions which seek to "win" with a division of payoffs.29 An analyst's voice in the
28Michael Handel, ''Avoiding Political and Technological Surprise in the 1980's," in Roy Godson ed.,
Intelligence Requirements for the 1980's: Analysis and Estimates (New Brunswick: National Strategy Information
Center, 1983) pp.103-104.
29The resulting payoff is primarily the credit received for providing accurate and useful interpretation of data.
28
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
A
group can thus be seen as his bargaining power. The desire to have part of his view expressed in
the final product may entice the analyst to "agree" with others on a different topic with the
expectation that they will reciprocate. Bargaining or "logrolling" may well inhibit the estimative
process.
The question of integrity in group analysis projects can be further aggravated by other aspects of
consensus-building. The momentum of consensus can pose a threat to the consideration of
alternative views. Once coalitions are formed it is very difficult to alter opinions, except in the case
of a radical change in events or the emergence of new information. The development of a
consensus also results in a tendency to stop looking for information that might contradict the
decision or to fail to report facts that are found which call the decision of the group into question.
Joint analysis situations also increase the costs of dissonance, resulting in the discouragement of
dissenting opinions. The incentive to dissent from the group opinion is further reduced by the
responsibility costs. In group decisions the personal responsibility any individual member has to
assume is very low. By raising a dissenting opinion, however, one forfeits his immunity from
blame, and has to assume sole responsibility for his statement.30 The pressures to coordinate
intelligence can reach a point at which alternative views or strong disagreements become highly
unlikely and thus legitimate differences are obscured.
Political Surprise
Although issue of surprise did not serve as a primary focus of this conference, we deal with it
briefly here as a general problem that analysts must confront. Political surprise consists of
unforseen domestic and diplomatic events that effect United States foreign policy. There are three
elements which contribute to surprise: 1) those that are self-generated, such as problems of mindset
and other flaws in the analytical process31; 2) those that are inherent in the nature of world events;
and 3) those that are caused by the enemy, known as deception; We discuss the latter two elements
of surprise here.
Loch Johnson, "Decision Costs in the Intelligence Cycle," in A. Maurer ed., Intelligence Policy and Process
(Boulder: Westview Press, 1985) pp.187-188.
30Loch Johnson, "Decision Costs in the Intelligence Cycle," in Intelligence Policy and Process pp188-189.
31We will discuss these problems in detail in a later section.
29
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
A
Solid analysis can often lay out a range of possibilities and even sometimes predict future
occurrences, but it will never eliminate surprise altogether. In this world of uncertainties, therefore,
intelligence efforts should concentrate on minimizing the effects of surprise in recognition that it
analysts will never eliminate it entirely. The most effective measures analysts can take against
political surprise are to increase their awareness of and sensitivity to U.S. national interests. By
understanding the areas of vulnerability to political events in other countries, the intelligence
community can lessen the impact of a surprise. The event is thus reduced "from a surprise in
subject and timing, to a surprise solely in terms of timing."32 One could avoid potential problems
with passing too many options on to policymakers by playing out the various scenarios only within
the intelligence community. This would ensure that when a crises occur, policymakers will receive
timely and well reasoned analysis without forcing them confront the implications of a range of
alternatives every time they read an estimate.
The complexity of the analytical task is further compounded by enemy efforts in deception that
cast doubts on some of the information with which an analyst must work. Dealing with deception is
made particularly difficult by the fact that the analyst often does not know for certain when he is
being deceived. The strategy of deception is usually to strengthen ideas that already exist in the
mind of the enemy and thus it is often tailored to the preconceptions of analysts. Another problem
in combating deception is that "the more alert one is to deception, the more likely he is to become its
vicdm."33 Thus overconcern with deception can on one hand paralyze an analyst and force him to
disregard valid information or, on the other hand, increase his susceptibility to certain deceptive
techniques.
On an individual basis, analysts can screen intelligence estimates for deception by laying out the
key judgments, assumptions and logic employed in their production and then estimate the extent to
which the estimate would change if any of these components proves to be wrong. Analysts can also
examine the information sources exploited in each component of the analysis to determine the
multiplicity of sources behind each assumption. There must also be a systematic determination of
the susceptibility of information to deception and the issues which need extra assurance against
32Ibid.
33Michael Handel, "Strategic Surprise: The Politics of Intelligence and the Management of Uncertainty," in
Intelligence Policy and Process, p.245.
30
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
deception. Although self-awareness cannot remove the threat of deception, it can minimize
deception's impact on the intelligence process, as well as improve the overall effectiveness of the
intelligence process by highlighting gaps in knowledge so that tasking is more focused.34
On an organizational level, there is significant debate over the potential effectiveness of a special
counter-deception branch within the intelligence community. The skeptical view of the prospects
for developing a counter-deception system is that the problem of deception is virtually unsolvable in
any systematic manner and thus should be left to the individual analyst to address in a personal
manner. Additionally, this general viewpoint maintains that there is no need for a special unit in the
analytic section to deal with deception, because the field of counterintelligence already exists. The
basic proposal for counteracting deception is simply skepticism.35 The large degree of
redundancy in the collection and analytical steps of the intelligence process serve as the best source
of error correction.36 In contrast to this position,
1)We believe that a counter-deception unit might effectively combat intelligence deception.
Such a unit would not eliminate all problems associated with deception, but it might reduce their
occurrence. The complex nature of the problems posed by deception require the formation of
special teams to focus exclusively on this issue. Such a team should be small and flexible and have
high security access. Participants on these teams should be limited to top level analysts perhaps in
combination with properly trained academic specialists and terms of service should be limited to a
few years. This system ensures that well qualified analysts will deal with the complex topic of
deception. At the same time, limiting the amount of time an analyst can spend on the task force
reduces the possibility that the group will fall prey to a mind-set concerning deception. Analysts
will then be able to return to their jobs with an even better perspective of the pertinent issues.
Problems on the Level of the Individual Analyst
Every analyst is limited by the personal, political, and cultural biases he maintains. These
34Ibid., pp.146-150.
35William Harris and Raymond Rocca, discussants of papers by Epstein and Feer, in Intelligence Requirements
for the 1980's : Analysis and Estimates, pp.152-157.
36From an interview with an analyst in INR.
31
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
analytical biases often prove particularly difficult to combat because individuals often do not
recognize their influence. Personal and political biases vary from person to person, so the multi-
layered bureaucracy of the intelligence community should help to cancel out these individual biases.
The issue of cultural bias is different, because it tends to be reinforced through the bureaucratic
process. Cultural biases arise from a lack of understanding of other cultures, which force the
analyst to view them through stereotyped images that prejudice and distort perception. For
example, the prevailing belief in 1957 that the Soviet Union lagged far behind the U.S. in science
and technology resulted in the opinion that it was incapable of launching a satellite, despite the
existence of much evidence to the contrary.37
Laqueur claims that cultural bias "is not true bias but parochialism, a deficiency of knowledge
and imagination."38 He describes the manifestation of bias in the intelligence community in terms
of the Cassandra and Pollyanna syndromes. The Cassandra approach is more commonly known as
worst-case analysis. It could lead to excessive defense spending, frequent alerts against possible
attacks, and rejection of peace feelers sent by an enemy.39 In this age of nuclear weapons, this
increase in tension is destabilizing and dangerous. Furthermore, frequent warnings increase
susceptibility to the "cry wolf' syndrome in which so many alerts are sounded, that in the case of a
real danger authorities no longer heed the warnings. Worst-case analysis provides an easy way to
escape responsibility, but reduces the quality of intelligence.40
The Pollyanna syndrome manifests itself as extreme optimism and overconfidence. This
strategy of intelligence is obviously very dangerous because it increases vulnerability to surprise.
The classic example of this type of analysis is the refusal of the Israelis to believe that the Arab
countries would attack in 1973. This understanding was the result of the overrating of the Israeli
superiority, and the underrating of Arab military strength, Egyptian ability to learn from previous
mistakes, and the Israeli capacity for deception.41 Best-case analysis is often the result of
37Michael Handel, "Avoiding Political and Technological Surprise in the 1980's," in Intelligence Requirements
for the 1980's: Analysis and Estimates, pp.90-92.
38Walter Laquer, A World of Secrets. The Uses and Limits of Intelligence, (New York: Basic Books, 1985),
p.280.
39Ibid.
40Michael Handel, "Strategic Surprise: Politics of Intelligence and Management of Uncertainty," in Intelligence
Policy and Process, pp.256-257.
41Walter Laquer, A World of Secrets, p281.
32
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
ignorance of the facts or simplistic reasoning, in which dimensions of events abroad that are critical
to an understanding of what is happening are not considered. The problem arises when significant
dimensions of the situation are ignored "such that each analyst touches one or two parts of the
problem but generates a mythical view of the whole by overgeneralizing from the parts."42
Another type of cultural bias is projection, in which analysts overrationalize events and project a
single personality or plan of action on a foreign govemment.43 Positive projection occurs when
an analyst perceives a similarity between him and the opponent and projects a fundamental similarity
in viewpoint.44 This appeared as the major cause of the underestimation of Soviet ICBM
deployments in the late 1960's and early 1970's according to a U.S. Senate report. CIA analysts
projected a similarity in views of arms control and deterrence held by the U.S. and the Soviet
Union. Since the Soviets did not have a similar perspective on the issue, U.S. intelligence estimates
were useless and misleading.45 An equally dangerous situation may arise in negative projection,
when an analyst does not perceive any similarities between his views and those of a foreign
country, so assumes they will always contradict U.S. opinion and behavior. Both types of
projection result from simplistic reasoning in which major elements of a country's culture and
political climate are ignored in the analytical process.
The final type of cultural bias, mind-set, can aggravate other problems of bias and
misperception, by the way in which it influences almost every step of the analysis process. An
analyst's preconceived notions determine what type of information will be collected. His attitude
about a particular source of information may influence the weight he ascribes to the data he receives.
It colors the way in which he perceives and analyzes the information, as well as what types of
information he decides to pass on to the next level in the bureaucracy. The problem of mindset in
analysis can be described as the attempt to make reality conform to pre-existing concepts instead of
deriving concepts from an observation of reality. The U.S. House Permanent Select Committee on
Intelligence cited mind-set as the major cause of the CIA's failure to predict the Iranian Revolution.
It stated that the underlying notion of the policy and intelligence communities was that the Shah
42James Austin,"The Psychological Dimension of Intelligence," in Intelligence Policy and Process, p.210.
43Robert Jervis, Perception and Misperception, pp.319-328.
44James Austin, "The Psychological Dimension of Intelligence Activities," in Intelligence Policy and Process,
p.210.
45U.S. Senate, Foreign and Military Intelligence, pp.76-78.
33
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
would be in power well into the 1980's, and that Iran was not likely to have a revolution. This
belief (or wishful thinking) was "supported" by the CIA's discouragement of the collection of
information that would demonstrate that Iran was on the verge of revolution.46
Problems in analysis caused by the biases or mind-sets of individuals can never be totally
eliminated. The effects of such biases can be reduced through the filtration of analyses through
many levels of an organization. Additionally, as argued above, acknowledgement of underlying
assumptions could decrease the susceptibility of intelligence analysis to the negative effects of bias.
The conference therefore proposes that
2) The training of analysts should emphasize an awareness of the problems of bias.
Evaluation of Analysis
The final issue specific to the analytical process is the evaluation of analysis. The question-How
can analysts learn from their past analytical work and improve on it in the future?- must serve as the
basis for discussion of the intelligence evaluation process.
Although the need for evaluations of analysis is universally accepted, the criteria by which
analysts will be critiqued is not. Some argue against an evaluation process in which an analyst
receives a "grade," because "judging yesterday's analysis based on today's knowledge would
provide a somewhat skewed view."47 Instead they proposes a system which would provide
analysts with "lessons to be learned" from the analytical exercise. Others agree that the mechanisms
to evaluate analysis cannot be based primarily on the accuracy of predictions, because such a system
would pressure analysts to produce estimates with short time frames, vague conclusions, or safe
topics. It is thus important to establish an evaluation system which is flexible, yet objective.48
The important distinction between the two frameworks is that the latter supports a graded system of
evaluation, modified to eliminate certain biases.
46U.S. House of Representatives, "Iran: Evaluation of U.S. Intelligence Performance Prior to November 1978,"
Staff Report Subcommittee, Permanent Select Committee on Intelligence, January, 1979,pp.2-7.
47Arthur Hulnick, "Managing the Intelligence Process: Strategies for Playing the End Game," 1987 Annual
Meeting of the American Political Science Association, 3-6 September 1987, p.16.
48Glenn Hastedt, "Organizational Foundations of Intelligence Failures," p.143.
34
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
There are two, not necessarily exclusive, possibilities for reviewing intelligence products: one
which uses people entirely within the intelligence organization and one that employs outside
experts.49 The first divides into two basic modes of evaluation, peer review and a devil's
advocate system. In the peer review mechanism, other analysts working on related topics use their
understanding of the issues to review and evaluate the intelligence products of their colleagues.
Analyses are evaluated for the thoroughness and soundness of the analytical processes. In the
devil's advocate system, a special group of analysts have the sole charge of challenging the
conclusions reached by their peers. Supporters of the devil's advocate system claim that this
evaluation mechanism reduces the problems caused by the biases or mind-sets of analysts.
Some argue that evaluations by experts outside the analytical community is essential to having
an objective system.50 Additionally, outside reviewers may more effectively adopt the
perspectives of the top level policymakers, and thus may better evaluate the effectiveness of the
intelligence product.51 On the other hand, outside reviewers have their own political and
institutional biases that may cause problems in their evaluation. Some feel that the Team Al Team B
exercises demonstrate this phenomenon.52 Another objection to outside evaluation is that people
outside the intelligence community may not understand the analytical process and thus may not
provide relevant criticisms and suggestions for improvements. Finally, outside review requires
access to sensitive information and some argue that people outside the intelligence community
should not be allowed such access.
3. The evaluation process should include a detailed review of the thought process of the
analyst.
It is essential that intelligence estimates are evaluated to provide an incentive for maintaining
high quality work and to ensure the existence of a channel for the flow of constructive criticism. In
order to avoid the problems associated with grading solely on the basis of accuracy, an evaluation
49The issues of Congressional and Executive oversight are discussed below.
50From a speech by Robert Gates, the Deputy Director of the CIA, delivered at Princeton University on
September 29,1987.
51From an interview with an intelligence community official.
52Arthur Hulnick, "Managing the Intelligence Analysis Process: Strategies for the End Game," p.11.
35
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
should consist of a detailed review of the entire analytical thought process. Such an evaluation
would take into consideration the nature of the analytical problem and the amount and quality of
information that was available to the analysts. Although this process is more costly than many
others, it provides the best assessment of the problems in analysis. Additionally, such in depth
evaluations may reduce the number off reviews necessary by providing a formal mechanism to
improve analytical work rather than one to simply grade it. Overall then, the benefits of such a
system should far outweigh the costs.
Peer review is an important step in the analytical process, for it serves a dual purpose. First, it
provides an evaluation of analyses done from the perspective of an "insider" who understands the
inner workings of the system. Second, it enables analysts to have a greater sense of other research
that is being conducted around them. This type of evaluation system is not complete, however,
because it can suffer from the presence of institutional biases or mind-sets. The devil's advocate
system would not alleviate these problems because the evaluators would also eventually suffer from
a fixed frame of reference and would eventually be dismissed as "crying wolf."
4. Important analytical work should be evaluated by a committee composed of both former and
active members of the intelligence community and outside experts.
Such a committee would ensure that a variety of backgrounds, perspectives, and interests were
brought to the review process. The analysts would see that the evaluation is performed in light of
the circumstances surrounding intelligence analysis while outside experts would provide fresh
perspectives on the issues at hand, unhampered by institutional biases. The issue of classification
and secrecy should not be a problem because some scholars have the type of security clearance
necessary to participate in the review process. Since this process will only be undertaken for the
most sensitive issues, the added assurance that the analysis is not tainted is worth the risk of
providing classified information to a few select people.
Personnel Quality
Many of the specific issues of analysis revolve around the quality of the analytical personnel.
36
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
The unavoidable human element of analysis points to gathering superior analysts as one of the most
important factors in producing high quality intelligence. In this next section, we explore the
recruitment, selection and training of analysts by the CIA in order to recommend measures to
improve personnel quality.
1) A significant increase in the number of recruiting offices and recruiters is warranted.
The primary problem with recruitment is that there are not enough recruiting offices or
recruiters. Ten recruiting offices and no more than forty recruiters are simply not enough to contact
the amount of people necessary to fill the CIA with excellent personnel. After World War Two, the
"old boy" network provided the CIA with high quality and experienced personnel. Recruiting was
done on a very personal basis. Recruiting today has become much more structured and
bureaucratic. The absence of old boy networks to provide the CIA with personnel necessitated such
a change; however, the bureaucratic structure has not consistently provided the CIA with as many of
the same quality of personnel as before. The CIA should dramatically increase its recruiting efforts
in order to ensure that the type of recruit that in the past made a successful CIA analyst once again
will consider the CIA as a career.
In addition, public opinion also has a clear and direct effect on the ability of the CIA to recruit
new employees as well as to attract academics in lateral employment exchange programs. A
negative public perception of intelligence and the CIA can have a devastating effect on recruiting
efforts. The negative effect on recruiting was demonstrated clearly in the mid-1970s, when public
criticism of covert intelligence activities caused applications to dwindle and forced the agency to hire
older, less qualified personnel. Today, the existence of anti-CIA factions on campus hampers
recruitment efforts and denies intelligence organizations a substantial number of qualified people.
Opinion towards the CIA is particularly hostile in academe; this hostility makes lateral exchange
programs unpopular and often unfeasible and bars intelligence access to important human resources.
Public relations recommendations are discussed in a later section.
37
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
2) The testing procedures need to be more comprehensive, ideally involving a writing sample
and an increased focus on foreign language skills..
The next phase in the employment process is selection of personnel. The selection process is
quite detailed and time-consuming taking from six to nine months to complete. It consists of eight
parts: a resume, an interview with a recruiter, the Professional Applicant Test Battery, an
application form, an interview in Washington, a background check, a medical and psychological
test, and finally a polygraph exam. This process focuses, as it must, on security issues, but often
without devoting adequate attention to candidate qualifications, notably analytical skills, language
ability, and writing ability.
The CIA adequately tests the candidates' ability to think in pressure situations on the
Professional Applicant Test Battery, but does not examine the methodical, long-term problem
solving abilities, critical to good analysis. Furthermore, the CIA requires no writing nor does it
place sufficient emphasis on foreign language capability upon entrance. The testing of foreign
language capabilities is somewhat superficial in that it only asks the applicants to rate their reading,
writing, and speaking abilities on a scale from one to five. Recognizing that both clear writing and
foreign language proficiency are critical to good analysis, the CIA should test these skills more
systematically.
3. The CIA should increase the amount of training given to analysts.
Analysts often receive an inadequate amount of training. While operators receive lengthy and
comprehensive training in the Career Training (CT) Program, which involves both classroom
instruction and practical application, analysts too often receive only on the job training. Analysis
may be weak due to the lack of specialized, intensive training of the analysts. For instance, entry
level training programs in analytical and historical basics, mid-career leaves of absence for study at
universities, and extended, and more specialized, in-house analytical training seminars might all
improve the capacity of the intelligence community's analytical personnel. Importantly, the
community must ensure that all analysts receive such training, both at the beginning and at intervals
38
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
throughout their careers.
4. The CIA should provide more supergrade positions for analysts and thereby increase the
number of high level people doing analytical work.
The best analysts are often sent to the CT Program to learn about the non-analytical branch of
the Agency. This enables them to become generalists and hence better managers for the CIA and
many follow career paths that move them into managerial positions and out of analysis. While the
CIA should certainly have some analytical people among its managers, it should also increase the
number of analysts at supergrade levels and thus establish a career path in which excellent analysts
could receive promotions and salary increases while still doing analytical work.
The intelligence community will never completely eliminate human error as a source of analytical
weakness. The effect of human fallibility, however, can be minimized by ensuring that the
analytical community consists of high quality personnel. The intelligence community needs to
place more emphasis on the personal aspects of analysis. Better recruitment, selection, and training
would contribute markedly to the quality of analytical intelligence.
39
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
The Producer/Consumer Nexus
Organizational changes within the intelligence community and reforms in the analytical process
itself will improve the intelligence product. In order to be effective, however, intelligence estimates
must contribute positively to the policymaking process. Intelligence work does not occur in a
vacuum. Its purpose is to provide policymakers with information and analyses upon which to base
their decisions. Therefore, any attempts to improve the intelligence community must account for the
effectiveness of the intelligence product in the formulation of foreign policy.
Often the fault for the ineffectiveness of the intelligence product lies with the intelligence
consumers themselves. For a variety of reasons, policymakers frequently disregard intelligence
analysis. They are often so busy that they cannot take the time to carefully read and digest
intelligence reporting. All too often therefore, consumers, especially at high levels, receive
incomplete versions of estimates in the form of either oral briefings or executive summaries
prepared by aides. They ignore long term issues which intelligence analysis could be well suited to
illuminate and often disregard intelligence reporting that would contradict their policy
preconceptions.53 They do not always know exactly what type of information or analysis they
need from intelligence and thus unintentionally miss useful estimates. Additionally, they sometimes
feel that they are superior analysts to those in the intelligence community and thus are reluctant to
rely on other's estimative work.54
In light of the observation that policymakers often neglect intelligence reporting, the intelligence
community should take steps to improve the producer/consumer interface. Awareness of
policymaker needs and concerns, appreciation for the limitations of intelligence and understanding
the problems inherent in the relationship between producers and consumers constitute a takeoff
point for examining the problems and proposing some suggestions.
In order to combat the problems associated with the producer/consumer relationship, we believe
that a combination of structural adjustments and attitudinal changes can make a difference. Our
53The history of the Vietnam War provides numerous examples of policymaker reluctance to hear negative
reporting.
54Henry Kissinger's term as National Security advisor and then Secretary of State provides the most vivid
example of this phenomenon.
40
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
general recommendations rest primarily on fostering closer contact between those who do analysis
and those who use it in making policy. Consideration of the question of distance suggests that the
benefits of policy/intelligence proximity outweigh the costs' of such a close relationship. Although
analysts must certainly minimize the effects of politicization on their work, most experts agree that
in practice "cooked intelligence" is a rare phenomenon (tales of former DCI William Casey
notwithstanding), while overemphasis on divorcing intelligence from policy cripples the ability of
intelligence to enhance decisionmaking. To be useful, intelligence must be relevant, timely, and
easily assimilated by the decisionmaker. Proximity in the relationship between analysts and
policymakers will enhance the contributions of intelligence to foreign policy decision making.
Having asserted that a closer relationship between intelligence producers and consumers is
desirable, we must design mechanisms to foster such a closer relationship. In order to do so we
must base our response on certain key observations of intelligence analysis' role in the
policymaking process.
1. Both communities should draw on the high degree of informal contact that already exists to
further increase producer/consumer contact.
We recognize that the foreign policy bureaucracy has developed an informal system of
communication and decision making that complements the official hierarchical order. We see
repeated examples of this phenomenon in the daily, informal contact among intelligence analysts and
between analysts and their counterparts in the policymaking community. Rather than proposing a
drastic alteration of the bureaucratic structure of the foreign policy community, we instead try to
emphasize the importance of this "natural system" to improve the degree of producer/consumer
contact. Consistent with our earlier discussions of organization and analysis, we recognize that
drastic bureaucratic changes may often prove counter-productive and thus should be instituted only
when there is a direct and pressing need.
41
L Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
2. Action should be taken to increase contact between middle level analysts and working level
intelligence consumers.
Certain types of decisions generally filter up through the foreign policy bureaucracy, yet no
formal structure exists to promote junior producer-junior consumer interaction. Working level
policymakers in the NSC staff or the State Department generally play a more active role in daily
policy regarding individual countries or issues than their superiors. Middle level policymakers also
provide background and frames of reference for the more important policy decisions and thus
ensure themselves of a certain degree of influence; indeed, senior officials often merely select their
choice from a policy menu prepared by their subordinates. These middle level officials,
furthermore, are more likely than senior officials to have time to study detailed intelligence reports.
The analytical community, should therefore target middle level officials for increased emphasis as
intelligence consumers in recognition of their importance in the policymaking process.
To foster greater contact between middle level intelligence consumers and middle level
intelligence producers, intelligence agencies should make an effort to match each mid-level
policymaker with intelligence counterparts at the appropriate level in the intelligence hierarchy. All
policymakers should have ready access to and personal familiarity with intelligence officers who
could provide the intelligence support demanded by the decisionmaker's agenda. This official
adoption of an informal tendency would improve the timeliness and relevance of much analytical
work.
Middle level contact could be further improved through formal briefings of middle level
intelligence consumers. Briefings require face-to-face contact, which would place the relationships
on more personal bases. This might have the added benefit of promoting greater informal contact
between producers and consumers. Policymakers might be more likely to call analysts they know
in the intelligence community for opinions than people with whom they have no personal
relationships. Finally, the intelligence community could promote greater consumer/producer contact
at middle level through workshops and informational seminars that bring together analysts and
policymakers working on the same issues. Such a program has met with some success already at
42
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
_Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
DIA. We recognize the strain that such seminars might put on already limited time, but we believe
that if indifference among analysts and policymakers to such programs can be overcome, they
would be worthwhile. Not only would these workshops directly increase information and analysis
sharing, but again, they might facilitate informal communication.
3. Contact between middle level analysts and high level policymakers should be increased.
Our final important observation is that middle level analysts, who concentrate all their work on a
specific country, region, or issue area, can often provide better assessments of the situations within
their particular specialties than their superiors. Upper level analysts and managers receive most of
their information second- or third-hand. They depend upon those who work for them to provide
either the analysis or at least the basis for the analysis that they confer to policymakers.
Accordingly, one should consider the importance of middle-tier analysts when attempting to
improve the quality of analysis passed to policymakers.
Although senior policymakers cannot be expected to devote as much time as their staff to
perusing detailed intelligence products, efforts should be made to increase the contact that upper
level policymakers have with intelligence officers analysts. Rather than interacting only with high
level managers of the intelligence community, senior officials could benefit from more contact with
the mid-level intelligence officers who are actually writing the intelligence products. Some senior
officials are reluctant to be briefed by lower level analysts, but such meetings have great potential to
increase policymaker awareness of important intelligence issues. Both President Kennedy and
President Carter worked effectively with junior intelligence officers, and other senior policy officials
would also benefit from such interaction. Along with middle level policymakers, senior
decisionmakers, should be encouraged to develop personal relationships with intelligence analysts
to facilitate policy-intelligence interaction.
4. There should be an increased emphasis on long term intelligence reporting.
The other primary problem in the use of intelligence to enhance foreign policy is a disturbing
43
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
lack of appreciation of long-term estimates by decisionmakers. "Preoccupation with current
reporting is, from our perspective," said DDCI Robert Gates, "a major problem." 55 The average
tenure of an assistant secretary in government is 21 months, which makes a short term focus
understandable, but still costly to the nation. Additionally, heavy time pressure often prevents
policymakers from thinking about issues beyond the immediate horizon. In examining intelligence
failure in the 1974 Cyprus crisis, the Senate Select Committee on Intelligence, too, decried the
proclivity of decisionmakers to focus on the latest raw data rather than on refined analysis. As
DDCI Gates laments, "For many years, trying to get senior policy principals to meetings to discuss
longer range intelligence requirements has been an exercise in frustration." 56
The focus of policymakers on short-term issues, with the concomitant pressure for production
of current intelligence studies, has damaged the intelligence community's long-term capabilities.
Analysts like to feel that their work is appreciated, so neglect for long-term work has made positions
in current intelligence more prestigious than positions in long-term estimating. Current work is
"more interesting, more exciting, and guarantees the analyst more visibility." Top analysts are
attracted to current intelligence at the detriment of long-term studies.
Moving policymakers to place increased value on long-term issues is a difficult task; one cannot
legislate change in decisionmakers' attitudes. One step to improve awareness of long-term issues,
however, is possible: meetings can be arranged at which intelligence officials who concentrate on
specific geographic regions brief policymakers on long-term trends and their likely consequences.
Institutionalizing such meetings at set time intervals should reduce the chance that policymakers
form decisions without considering long-term trends.
Much of the weakness in the use of intelligence to enhance foreign policy is a result of the
attitudes of intelligence consumers. Improvements in the quality and packaging of intelligence
products are immaterial if policymaker never reads the intelligence report. Changing attitudes,
however, is a difficult task; one cannot simply legislate that decisionmakers must place greater
weight on intelligence products and long-term estimates when forming their opinions. Although
55Robert M. Gates, "CIA and the Making of American Foreign Policy," public address at the Woodrow Wilson
School of Public and International Affairs, Princeton University, September 29, 1987, p.20.
56Ibid. p.19.
44
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
a altering consumer attitudes is a daunting challenge, such modification is necessary, nevertheless.
To eliminate producer/consumer misperceptions and improve the quality of intelligence tasking,
the level of policy-intelligence officer interaction must be increased. Increasing the dialogue
between the intelligence producer and consumer should improve the relevance of the intelligence
product to the decisionmaker. Making intelligence products more useful to consumers should
increase the proclivity of the decisionmaker to consider intelligence reports when making policy. In
this manner, increasing interaction at the producer-consumer nexus should enhance the contribution
of intelligence to the formation of foreign policy.
45
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Covert Action.
Though clandestine operations account for only a small fraction of the resources of the intelligence
community, their very nature assures that they account for a disproportionately large percentage of
public and governmental attention. We approach the question of covert action from a variety of
angles, considering structural changes that have been suggested as well as longer term trends
associated with clandestine activity. Despite its small financial cost relative to the rest of intelligence,
covert activity will always be a highly visible and controversial subject, and the conference has treated
it accordingly.
Since World War Two, American public opinion towards intelligence and intelligence
organizations has vacillated, tending to parallel swings in national foreign policy consensus. In times
of national foreign policy consensus, public support for intelligence has increased. This was case
during the Cold War of the 1940s and 1950s, for example, when exposure of covert intelligence
operations engendered little public criticism. Recently, national consensus support has forestalled
negative publicity about covert action in Afghanistan.
Where national consensus dissolves in the face of dissent among Congress, the executive, the
media, and the public, public support of intelligence has tended to decline. The 1960s and 1970s,
periods of breakdown in American Cold War consensus, witnessed a sharp decline in public support
of foreign policy, covert action, and intelligence operations in general.
Today, unqualified foreign policy consensus no longer exists; media and public polling indicators
reflect an ambivalent public attitude towards specific policy decisions, among them covert operations.
In an environment that will challenge policy choices, policy makers will have to develop intelligence
and covert action policies that will be consistent with overt policy declarations; as the Iran-Contra
affair demonstrated. Failure to do so will engender devastating public criticism and distrust.
46
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Implicit in our analysis, therefore, is recognition of the fact that in modern America, covert
actions subject to scrutiny and judged by the public on their merits alone. Though just about everyone
agrees that the United States should engage in some forms of covert action some of the time,
policymakers no longer have carte blanche. The public is not willing to condemn covert operations
wholesale, nor is it willing to accept them blindly. The advisability of any particular covert action,
thus, becomes a question of judgement. The task of the policy maker is to limit, authorize, and
regulate operations so as to mandate processes which assure that careful judgement will occur. This
can be done by establishing implicit rules, by congressional and independent oversight, by internal
agency directives, by executive order, and by Congressional statutes. The conference proposes the
following recommendations
I. The Central Intelligence Agency should design its covert operations with the assumption that
they will be made public.
Robert Gates, DDCI, said in a speech at Princeton University 57 that the American people have,
"a contradictory melange of images of CIA and very little understanding of our real role in American
government." The public has a perception of the CIA different from any other government agency,
bureau, or department, mostly because many CIA activities are secret. Secrecy leads to speculation,
and speculation leads to suspicion. The very gain of American ideals and institutions runs against
government secrecy. Though considering a specific issue, the Church Committee provides an
important example of the contention that intelligence activities should be open to public scrutiny:
The truth about assassination allegations should be told because
democracy depends upon a well-informed electorate. We reject any
contention that the facts disclosed in this report should be kept secret
because they are embarrassing to the United States. Despite the
temporary injury to our national reputation, the Committee believes that
foreign peoples will, upon sober reflection, respect the United States
more for keeping faith with its democratic ideal than they will condemn
us for the misconduct revealed. We doubt that any other counts), would
have the courage to make such disclosures.58
57This speech served as the source for his article found in Foreign Affairs, winter 1987
47
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
A government agency shrouded in secrecy is bound to incur much conspiratorial theorizing. We
have indeed found that bureaucracies often hide behind a wall of excessive secrecy and classification.
59 The knowledge that one's designs will never be open to scrutiny makes for bad policy. Secret
power is easily abused, as former DCI Stansfield Turner writes: "Secret, unaccountable power is
subject to misuse, ranging from deliberate, improper diversion of resources to just plain carelessness
in the making of decisions."60
The public shares these deep suspicions. Taking media coverage as an indicator of public
attitudes, we find that from 1974 to 1978, of the 1,079 intelligence stories aired on network news
broadcasts, 714 (68.2%) portrayed intelligence unfavorably, 141 (13.9%) portrayed intelligence
favorably, and 224 (17.9%) portrayed intelligence neutrally. 61 Additionally, a review of long term
polling data reveals a slow but steady decline in the prestige and public rating of intelligence
organizations. Despite the renewed consensus and foreign policy optimism in the early Reagan years,
public opinion towards intelligence did not rally as bouyantly as might have been expected: six
different series of questions presented by the Gallup and Roper polls between 1978 and 1986
indicated that widespread public skepticism still exists.
It is possible that no ethical model or principles can adequately resolve the tensions of a secret
institution operating in a democratic society. Stansfield Turner's practical solution is to convince the
public to trust the agency by convincing the public that the CIA will use serious judgement and will
accept responsibility for its decisions. His proposal is in the form of a judgement test that would
58Senate. Select Committee to Study Governmental Operations with Respect to Intelligence Activities.
Final Report. S. Rept. 94-755, 94 Cong. 2 sess. (1976) Bold( II; Intelligence Activities and the Rights of
Americans. 2.
59 The NSC's Information Security Agency concluded in a 1982 study that of the 16 million pieces of
infromation classified in 1980, 800,000 were classified unnecessarily, 600,000 were classified by
individuals wihout authority to classify, and 2.2 million by people who went beyond their authorization,
80 Stansfield Turner. Secrecy and Democracy. The C.I.A. in Transition. (New York: Harper & Row,
1985), 4
61Ernest W. Leferver and Roy Godson eds., The CIA and the American Ethic: An Unfinished Debate.
(Washington D.C.: Ethics and Public Policy Center, 1979.), 109.
48
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
reassure the public that a 'spirit of openness' exists behind the CIA's closed doors, as he told John
Ranelagh: "There is one overall test of the ethics of human intelligence activities. That is whether
those approving them feel they could defend their decisions before the public if the actions became
public."62 Yet another angle from which this tension was approached was that of William Colby,
who postulated the need for "honorable men" in the intelligence community, whose responsibility and
integrity could help guarantee public trust.
We agree with Mr. Turner's reasoning and propose that a light of day standard ('that all actions
should be designed to go public') would reassure the public that their opinion is taken seriously in
secret operations. Additionally, public accountability might help ensure that covert actions were
designed to be consistent with overtly accepted objectives.63 A light of day standard might be
institutionalized be requiring the Deputy Director of Operations, the DDCI, and the DCI to sign off on
descriptions of all covert operations as well as to sign off on periodic up-dates of the operations.
Though it is clearly undesirable for the majority of clandestine actions to actually go public, design
and quality might very well be improved if a light of day standard were adopted.
2. We recommend that accountability be increased, specifically at the higher levels.
Vertical responsibility is an example of a process which encourages propriety and good
judgement. Vertical responsibility means that accountability will move upward so that an operative in
the field as well as all of his superiors are held responsible for his actions. It encourages propriety
and good judgement because everyone in a position to be aware of the operation will know that he/she
is also responsible for that operation.
The Ford Administration instituted processes for vertical responsibility in the early 1970's,
62John Ranelagh. The Agency: The Rise and Decline of the C.I.A. (New York: Simon and Schuster,
1987.), 643.
63The light of day standard might, for example, have helped to forestall the recent Iran arms for hostages
scandal.
49
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
processes that were retained by President Carter. The Reagan Administration has, however, removed
them. According to at least one agency official, the results of this change have been poor. When a
CIA published Contra manual was discovered to be recommending methods of 'neutralizing' local
officials, the result in the agency was that three low-level officers were reprimanded. The chain of
responsibility moved directly downward. This type of accountability can hardly be expected to
encourage responsibility within the agency. The message which the Reagan Administration is
sending out to intelligence officials is: do not worry, if an operation runs afoul of public opinion or
moral standards, neither you nor the agency will pay the the price. The problem of accountability is
one that will surely have to be addressed early in the next administration.
To ensure responsible direction of covert operations, accountability must rise vertically upwards
from the agency-level, beginning with agency directors and deputy directors, and continuing on to the
DCI, the NSC and the President. At each step of approval, the signatuer must assume responsibility
for the sensitive activity.
In past administrations this upward-moving vertical accountability has worked well to guarantee
proper consideration of all risky proposals, and it was a primary factor in President Carter's desire to
halt covert operations. Under Reagan however, stepped-up secrecy and aggressiveness have
redirected vertical accountability downward. The Iran-contra affair provides a case in point, with CIA
seniors Casey and Gates refusing to approve or disapprove of the operation (once they found out
about it), forcing complete responsibility almost entirely on NSC staffer and the major operator Lt.
Col. Oliver North. 64 Accountability should have gone all the way to the President, but as the Tower
Commission found, "there is no evidence that (CIA) Director Casey made clear to the President that
64 After learning of the possible diversion of funds to the contras, Casey and Gates scheduled a lunch
with North to ascertain any CIA involvement in the operation: "[N]either CIA official w[as] interested in
hearing any more about it. All they wanted to know was whether the CIA was 'clean.' Assured by North
that it was, Casey and Gates pressed no further and again made no mention to any oversight body." See
James Bamford, "An Aspirin for the CIA, but Major Surgery Needed," The Los Angeles Times, 8 February
1987.
50
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
? Lieutenant Colonel North, rather than the CIA, was running the operation."65 Presidents and other
high-level officials are subject to backlash and public disfavor for failed operations occurring under
their jurisdiction regardless of whether they have given approval, one more of the many reasons why
they must be more directly involved in approving those operations.
3. More attention should be paid to the hiring and training of Clandestine Operations personnel.
Implicit rules which encourage propriety and good judgement are of paramount importance
because legal and institutional restrictions cannot outlaw 'unethical behavior'. It must be recognized
that even a dramatic increase in the number of rules and laws restricting operatives could not prevent
every possible impropriety. In the final analysis, the man in the field will often act in the way that he
or his immediate superior sees fit. There can be no strict guidelines which cover all ethical concerns
and all international situations. More importantly, the issues involved are ones on which intelligent
people may differ according to their world views. There can probably be no a priori rules to solve the
dilemmas of covert operations. The advisability of any particular covert action becomes a question of
judgement.
Judgement in itself, however, is a function of the character of the operatives. Because operatives
necessarily have a good deal of discretion due to the nature of covert operations, ensuring that these
operatives be men of good judgement as well as of good character should be a first priority of the
agency. In an earlier section, we discussed possible personnel and training changes and call to the
attention of the reader that in the area of covert operations, where the operatives are vested with a great
deal of power, these concerns must receive "heightened scrutiny".
4. We urge that the use of third parties in U.S. operations be strongly discouraged.
65 Statement of Edmund S. Muskie, summarizing the findings of the Tower Commission. See
President's Special Review Board, The Tower Commission Report, ed. The New York Times (New York:
Bantam Books, Inc. and Times Books, Inc., 1987), xix.
51
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
The Tower Commission reported that the United States relied heavily upon parties outside of its
control in the Iran-Contra Affair operations: "He (North) relied heavily on private U.S. citizens and
foreigners to carry out key operational tasks." 66 The commission recommended that usage of such
intermediaries be curtailed:
We recommend against having implementation and policy oversight
dominated by intermediaries. We do not recommend barring limited use of
private individuals to assist in United States diplomatic initiatives or in
covert activities. We caution against use of such people except in very
limited ways and under close observation and supervision. 67
There are two primary harms from working with third parties. The first is that they are often
individuals, groups or regimes with whom the United States would not wish to publicly associate in
secret endevours. The authors of The Iran Contra Connection write that the United States, "did not
hesitate to invoke the aid of neofascist foreigners." 68 Such national associations are troubling,
though in some cases, perhaps necessary. It seems reasonable to conclude that such cooperation and
association with 'improper' intermediaries should occur only when it can be controlled and when no
other alternative will maintain operational viability.
The second harm is that working with third parties often involves training and equipping them,
and the fact that this training and equipment is often later turned to ill-purpose. Marshall et al. argue
that U.S. support for Cubans battling Castro in the 1960's eventually led to the Latin American drug
connection of the 1970's and 1980's:
America's drug problem today is arguably, in large measure, an outgrowth
of the 'secret war' against Fidel Castro.... They (Cubans) had the
clandestine skills, the Latin connections, the political motivation, and the
requisite lack of scruples to become champion traffickers. 69
This claim is supported by U.S. law enforcement statistics: no fewer than 8% of the Bay of Pigs
66John Tower, Edmund Muskie, Brent Scowcroft. The Tower Commission Report. (New York: Bantam
Books and Times Books, 1987), 55.
679 Ibid, 98.
68Johnathan Marshall, Peter Dale Scott, and Jane Hunter. The Iran-Contra Connection. (Boston, Mass.:
South End Press, 1987), 20.
69Marshall et al, 134
52
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
veterans have been arrested and convicted of drug crimes. 70 Similar concerns about the outcome of
the Contra war are present today. Since 1985, it has been known that the Contras and Contra-
supporters are heavily involved in the U.S. narcotics trade in order to earn cash to purchase weapons.
1 In 1985, the Associated Press carried a story on the largest cocaine seizure in West coast history; the
convicted smuggler said that the profits were meant to buy contra arms. 71
There are some in government who feel that as long as a course of action achieves our ends, it
should be used. Senator Malcolm Wallop writes that "there are some people who can only be moved
by the lowest motives. It is not ethically impermissible to deal with such such people on their own
level since they are already corrupt. Their corruption might as well be put to the service of good
ends." 72 Though the conference takes no direct stand on the normative dimension of such a policy,
we have found that the negative political consequences arising from the use of such third parties have
been quite significant.
5. We recommend that clandestine operations remain vested only in the Central Intelligence
Agency.
Under current Executive Order, covert activity lies solely within the jurisdiction of the CIA.
Although the CIA has historically assumed this role, many studies have examined whether such
placement is optimal for U.S. foreign policy interests. Does juxtaposition of the advisory and action
arms within one institutional framework compromise the efficacy of either or both? Are viable
reforms to ameliorate the situation extant?
Those advocating significant structural changes and proponents of the status quo hotly debate the
necessity and effectiveness of reform; the former call for the removal of covert action capability from
70Newsday, The Heroin Trail. (London, Souvenir Press, 1975), 169. As cited in Marshall et al. 127
71Marashall et al., 79
72Malcolm Wallop. "Ethics and Intelligence," in The Military Intelligence Community. ed. by Gerald
W. Hopple and Bruce W. Watson. (Boulder, Colorado: Westview Press, 1986), 214
53
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
the CIA while the latter urge structural reform within the current framework. Those urging relocation
contend that covert actions have undermined public opinion about the CIA; this unfavorable image
impacts negatively on U.S. intelligence by dissuading many academics from offering consulting
advice to the CIA and damages recruitment, thereby depriving the CIA of valuable inputs. They claim
that removing covert actions from the CIA would cause this image to improve, thus encouraging an
increase in quality input. Opponents argue that merely removing the covert action function and
renaming it would be a superficial change that would not alter public opinion. Moreover, any
restructuring of this magnitude should be considered in light of how it will affect the U.S.
government as a whole, not just one of its constituent parts. Another reason often given for the
separation of covert action from the CIA is that under the present system both covert action and
standard espionage suffer. Both work out of the same overseas network and often use the same
operatives; thus a failure on the part of either collection or operation places both at great risk.
Additionally, the allure of covert action is such that it impinges on the quality of intelligence collection;
the emphasis is on the short-term, tactical information needed for the successful implementation of a
covert action rather than on long-term, strategic information.
The most common call for reform suggests that paramilitary covert actions be assigned to the
DOD. Paramilitary covert actions have a particularly high noise level, and they require tactical logistic
support. Thus, many argue that the military should undertake such operations. Such military groups
as the U.S. Army's Special Forces and the Rangers are cited as units where such responsibilities may
be lodged. Many also suggest that the Department of State have a larger hand in the review and
approval process because although they presently play some role, their influence is not considered to
be significant. Such a reform would allow the CIA to concentrate on collection and analysis, strip the
CIA of its policy-making and -executing role, place covert activity under strict military discipline, and
better correlate covert actions and foreign policy. Opponents to such a change argue that the
restrictions of the War Powers Act would compromise operational security. Additionally, many argue
that the military is not fit to assume control over covert actions. The caliber of military personnel is
54
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
too low, the military lacks the necessary political sensitivity, and the military is a large, unwieldy
organism incapable of mounting an effective small operation.
Proponents of the status quo argue that radical reform as outlined above would be deleterious to
U.S. interests, because it would emasculate the CIA and because clandestine collection and covert
activity are inextricably linked. Splintering the CIA would diminish its power base and thus damage
its stature in Washington. Moreover, the DCI's position as the principal intelligence advisor to the
President would be reduced sharply. Splintering the Agency would necessitate the reduplication of
the CIA's current clandestine network, an undertaking that would lower efficiency, raise costs, and
increase security problems. Moreover, such a split would directly contradict the historical lessons of
dysfunctional inter-bureaucracy rivalry; both the U.S., through the OPC-OSO wrangling of the early
years of the CIA, and the British, through the wartime experience of SIS-SOE conflict, have
discovered the dangers of inter-bureaucracy rivalry. In light of these harms, the retention of
clandestine operations within the CIA seems warranted.
There are, however, reforms that should be made within the current system. Recently, efforts
have been made for such improvements. Under DCI Turner, plans for covert actions were routed to
NIO's for their critique prior to final approval. Later, DDCI Inman sought to further
interdepartmental communication between the DDI and the DDO. In addition, he sought to restore
public confidence in the CIA. In the final analysis, the key to successful intelligence-policy
integration is communication. Historically, dividing responsibility for covert activity have led to
crippling communication breakdowns. Thus, efforts should be made to promote intra-agency
communication, as Turner and Inman did. One concrete proposal would be to have the Covert Action
Staff of the DDO submit proposals for covert actions to the relevant regional bureaus in the DDI and
in the State Department. Such reviews could be performed in a similar manner to those done by the
NIO's.
55
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Long Term and Regional Issues.
Traditionally, analyses of covert action have concentrated on the effect of the secret operation on
the specific country in which the action took place. Such a view, however, ignores the essential
nature of these major undertakings. Ultimately, their effects are not specific. Covert action does not
occur in a vacuum. Rather, it is deployed in a complex world where it is nearly impossible to hide the
events of one day from the next. This escalating complexity engenders unpredictable relationships
between events and their catalysts and an incident may affect policy 20 years after the event has
actually occurred. It is important to establish how well Washington's policy toward a specific nation
fared - whether or not that policy included covert action. But it is even more significant for
policymakers to determine how one policy can affect U.S. relations with the whole region or even the
world itself.
It is customary for those who support covert action to rest their beliefs on the assumption that
stability and safeguarding United States interests outweigh any consideration of long-term
ramifications. ,Such a view might on its face seem to be more practical than that of someone who
eschewed interventionism because of a belief in the sanctity of national sovereignty. Proponents of
covert action are concerned with what they perceive to be real danger; in such an environment moral
arguments have little power, they reason.
Based on our analysis, we challenge this belief in the practical nature of covert action. Though
there are important ethical dimensions to interventionism, we leave those aside in the present context.
Rather, we choose to confront the essential argument of those who favor covert action. An analysis
of past major covert actions has led us to conclude that the harmful ramifications of these operations
are so significant that they often render the operations impractical and counterproductive. While the
immediate benefit of a covert action may be stability and enhancement of U.S. interests, too often the
long-term effect of the operation renders this stability insignificant in light of its cost.
Despite often negative consequences, policymakers have continually relied upon covert action.
Every U.S. president since World War II has utilized secret operations to influence other nations.
56
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
And every president in that time has seen his own or one of his predecessor's covert action policy
destroy foreign policy initiatives. Covert action, we find, corrodes foreign confidence and strains
international relations; contradicts and confuses foreign policy; fosters domestic discord and
dissatisfaction; damages national reputation and diplomatic credibility; and often entangles the country
in prolonged civil wars abroad that may unintentionally escalate into unwanted military commitments.
An examination of past covert actions and the policies which led to their implementation reveals
that bureaucrats are often swayed by the short-term benefits of the secret operations; they regard
future ramifications as less consequential in the face of a pressing foreign policy dilemma.
Policymakers are often hindered by their desire to rid themselves of a difficult problem - such as a
Pro-Soviet regime - and therefore fail to examine the long-term results of their policy. This myopia
may stem from the natural political attitude of an administration which controls the White House for
only four years at a time. Nevertheless, such shortsightedness can become a curse for future
presidents and for the national interest, regardless of whether the covert action succeeds in the
immediate sense.
But American interventionist behavior is more deeply rooted; it is not just the desire for immediate
success which spurs on policymakers. Rather, it is also our tendency to view Marxism in any form
as a direct threat. Thwarting Soviet aggression at all costs - such sentiment has periodically overtaken
Washington and then been crafted into a foreign policy whose singular purpose is to prevent
communism form gaining influence in nearby or allied countries. Intervention becomes such a
policy's ultimate manifestation, a panacea for frustrated policymakers who scoff at the idea that the
United States must forego its enormous advantage in strength and sit down at the bargaining table.
The United States' obsession with anti-communism often leads to both covert and overt policies
which are designed either to support what may be unpopular right-wing regimes or to subvert left-
leaning leaderships. But for such policymakers, even more important than preventing communism is
maintaining order. This behavior results in the perpetuating of political stereotypes which enforce a
57
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
policy which makes it difficult for Washington to accept the fact that revolutionary governments in
certain countries may not necessarily be hostile to American foreign policy. The anti-communist bias
of American policymakers has often poisoned political judgement, thereby obfuscating any sense of
objective practicality. In reporting American incursions and North Vietnamese invasions in
Cambodia, for example, Henry Kissinger would always characterize Communist attacks as "deep"
and U.S.-South Vietnamese offensives as "shallow," despite their frequent similarities in extent.73
We find that policymakers have a tendency to become addicted to covert action, thereby poisoning
any desire to attempt peaceful solutions. Indeed, one of the most troubling conclusions we are forced
to draw is that there is snowball effect operating in the realm of covert action. Once a covert action
succeeds, the threshold for intervention is lowered, and policymakers often attempt to use intervention
with increasing frequency and decreasing compunction. For example, initial short-run success in the
Philippines amidst the Cold War obscured political pragmatism and encouraged policymakers to trust
covert action as an instrument of unconventional diplomacy. Seduced by the apparent omnipotence of
these secret operations, the United States would court danger by repeating its interventionary tactics
throughout Southeast Asia, all the while enlarging both its covert commitment and official
involvement.
In fact, by the end of the Eisenhower administration, the U.S.'s reputation and American foreign
policy would become so inextricable enmeshed with its global covert interventions that virtually every
military mishap, political upheaval, and social calamity, regardless of the true extent of U.S.
responsibility, would be attributed to the manipulative behing-the-scenes diplomacy of the CIA,
which, according to President Truman, had become "so removed from its intended role (of
intelligence collection) that it (was) being interpreted as a symbol of sinister and mysterious foreign
intrigue - and a subject for Cold War enemy propaganda." Given the scope, scale and intensity of
Cold War covert operations in Southeast Asia, for example, "there was hardly a single country in
73 William Shawcross, Sideshow: Kissinger, Nixon and the Destruction of Cambodia. New York: Simon
and Schuster, 1981, p. 416.
58
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
which it was not widely believed that the CIA had been behind some major event."74
Similarly, we find from an analysis of three covert actions - Iran in 1953, Guatemala in 1954 and
Cuba in 1961- that escalation is the norm, rather than the exception. The policymakers who
formulated strategy in Guatemala in 1954 had been both stunned and pleased by how seemingly easy
the Iran operation had been the year before. Their judgment was warped, just as the judgment of those
who decided Cuban policy in 1961 was skewed, by earlier covert action successes.
Heightened sensitivity to the future effects of covert policy is necessary. The unwillingness to
consider the long term as a time frame of importance is a dangerously self-defeating institutional
problem that continues to plague Washington. Thus, the major policy recommendation engendered
by the previous analysis would seem to be obvious:
6) Policymakers should be extremely wary of employing covert action, for the future
ramifications of large-scale military secret operations can and often are extremely serious. Covert
action is not practical; its short-term effects can vary, as demonstrated by the operations in Cuba and
Guatemala, but its overall ramifications are usually quite damaging to the interests of the United
States.
In light of this conclusion, the response of many would be an incredulous and outraged "Why?"
"Did we not insure 25 years of stability in Iran?" they would ask. "Did we not stop the threat of
communism in our hemisphere when we operated in Guatemala?" One cannot deny these facts. Yet,
at what cost were these perceived gains made?
In Iran, there is a regime which is fanatically anti-American, a government which threatens to turn
the Persian Gulf into a raging battleground. What would have happened had we attempted to work
with, rather than against the Mossadegh government? Could Washington have possibly assured
stability in a different, more peaceful manner? When policymakers have covert action as a fail-safe,
74Roger Hilsman, To Move a Nation: The Politics of Foreign Policy in the Administration of John F.
Kennedy, New York: Dell Publishing Co., 1967, p.86-87
59
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
they pursue diplomacy halfheartedly. We attempted to invaded Cuba less than half a decade after
that country's revolution. Diplomacy is rarely successful in such a short time.
In fact, our attitude toward more diplomatic options shed light on the essential nature of covert
action: it can poison rational, diplomatic policymaking. Its aura skewed Eisenhower's and Kennedy's
viewpoint toward Castro and the Bay of Pigs and hardened Ronald Reagan's position on Nicaragua.
Covert action is so tantalizing to policymakers that they are blinded to its fundamental nature. What
seems like a sterile operation to rescue an often failed policy is in reality blatant interventionism, an
attempt to engineer the destiny of a nation whose culture policymakers do not understand. Such
behavior is not ignored by foreign leaders or foreign masses.
Rather, it is reflected in an increase in both anti-Americanism and political polarization. Moderate
left-wing regimes feel compelled to seek assistance from the Soviet Union because they fear United
States meddling. Conservative military dictatorships believe themselves safe from intervention as
long as Washington acts only on its anti-communist urges and regions as a whole become alienated
from American diplomatic, economic and political overtures
Additionally, covert action can have the effect of confusing our allies and those whom we are
attempting to influence so much as to be extremely counterproductive. Though we often have some
confusion as our goal, it has often happened that countries have been so unsure as to which of our
policies signalled our real intention, that they were rendered unable to discriminate the U.S.'s true
desire and act accordingly. 75
There are, perhaps, situations where the United States government is truly left with no alternative
but to attempt a clandestine operation -- to free American citizens or property, for example, when all
other overtures have proved unsuccessful. It is the contention of this conference, however, that
covert action is resorted to with far greater frequency than is necessary because it is often viewed as a
simple and effective way to resolve complicated dilemmas -- removing a troublesome popular leader is
75 Prince Sihanouk in Cambodia in the 1950's and Laos's Prince Phouma are two natable examples.
60
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
in many ways easier than having to deal with his regime. Unfortunately, covert action is neither
simple nor is it effective in most case. Consciousness must be raised about the deleterious
consequences as well about the tendency for covert action to engender further covert action. High
standards for the design operation as well as high criteria for triggering the need for an operation must
be taken more seriously if we are to work our way out of the negative consequences that have resulted
from 40 years of an active covert policy.
61
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
OVERSIGHT
While the executive agencies and the Congressional committees responsible for oversight of the
intelligence community have a mandate to review all aspects of their performance, it is clear that the
oversight task concentrates primarily on covert actions. We have structured our analysis of oversight
according to the agency or branch responsible, considering Executive oversight first, then
Congressional, then the effect of public opinion on the intelligence community.
Executive Oversight.
Following the intelligence scandals of the mid-1970s, which resulted in the formation of
permanent intelligence committees in both houses, congressional involvement in oversight of the
intelligence community was significantly increased. Indeed, some officials today claim that
congressional influence on the oversight process is now equal to that of the executive branch. While
this conference recognizes the benefits of an intelligence community that is accountable to the
American public through its representatives in Congress, we also realize the advantages and practical
necessity of having the executive branch shoulder the majority of the oversight burden. Because
intelligence is a tool used by the President to advance foreign policy and national security objectives,
the most accountable and effective oversight elements will be those that answer to the Chief Executive
and operate within the executive branch.
One must note that with the creation of new supervisory bodies, such as the Intelligence Oversight
Board and the Office of Intelligence Policy and Review (housed in the Department of Justice), there
are more executive oversight elements today than ever before. These elements possess the mandates
and provide the potential for quite effective intelligence supervision, yet numerous deficiencies in
good executive oversight continue, as evidenced recently by the Iran-Contra affair. It is not the
mechanisms of oversight in the executive branch which are faulty, but rather the uses to which they
62
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
are or are not put. A central factor in this is the direction and stimulation they are given by Presidents
and Directors of Central Intelligence.
Nonetheless, it must be realized that not all failures of intelligence oversight are avoidable.
Because of the secrecy, politicization and the necessary risks which intelligence activities must operate
under, some oversight decisions are bound to go astray. Even if a project is subjected to the utmost
consideration before being legitimately approved, improprieties or illegal actions may still occur at the
operational end with the oversight elements being held accountable. Though some oversight failures
must be anticipated and accepted, a significant number of them would be prevented if the oversight
elements in place in the executive branch received the direct stimulation and guidance they need to
fulfill their mandates. Such mandates include ensuring the legality, propriety and quality of
intelligence activities, their concordance with foreign policy objectives, their legitimate approval and
review, and their proper guidance and supervision at the agency level. An inability to fulfill these
mandates has continually plagued executive oversight in recent years.
In the Reagan administration, the source of this inability to fulfill oversight mandates can be
directly traced to the President's general antipathy towards oversight and his consequent weakening of
the established supervisory bodies. Although President Reagan reinstated the President's Foreign
Intelligence Advisory Board (PFIAB) early in his administration, he effectively rendered this body
and the other independent advisory board, the Intelligence Oversight Board (I0B), impotent by
viewing the appointment of new members more as a chance to reward for campaign services than a
chance to ensure good oversight of U.S. intelligence. The new appointees, mostly wealthy
contributors or longtime friends of the President, lacked the stature and strength to subject the
intelligence community to the close scrutiny necessary for competent oversight.
Further weakening of the vitally important oversight boards occurred through lack of stimulation
and guidance from President Reagan and from Director of Central Intelligence William Casey.
Reagan removed the oversight of "propriety" in intelligence activities from the IOB's tasking
(previously the JOB had been charged with overseeing the "legality and propriety" of intelligence
63
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
activities) and did not directly task the advisory boards to investigate specific intelligence matters
(such as the CIA's involvement in the Iran-contra affair). DCI Casey circumvented the boards
whenever possible, failing to bring questionable activities, such as the possible diversion of Iranian
arms sale profits to the Contras, to their attention. Casey further prevented the JOB from fulfilling its
mission by moving the CIA's General Counsel, who kept the IOB informed of potentially
questionable activities, to a new office four blocks away from CIA headquarters at Langley, thereby
isolating him from many of his sources of information.
While actions such as these in the Reagan administration have prevented the mandates of executive
oversight elements from being fulfilled, these mandates and the structures through which to carry
them out do exist, and would, if properly utilized, provide adequate oversight of the intelligence
community. In an effort to further the goals of these mandates and hence the competency of executive
oversight, this conference recommends the following:
1. The President's Foreign Intelligence Advisory Board (PFIAB) and the Intelligence Advisory
Board (1.0B) should be strengthened.
In order to receive an adequate review and assessment of intelligence functioning, the President
must appoint enthusiastic, experienced members to the PFIAB who will take their investigations
seriously and provide independent recommendations. The President must further stimulate this board
by directly tasking it to investigate and advise on certain key intelligence issues and by ensuring that it
meets with the full cooperation of the intelligence community. Meetings of the board, currently
bimonthly, should be increased. Though we cannot envision any specific criteria that would ensure
the quality and effectiveness of the board members, it must be recognized that this is one area where
the quality of the personnel is of paramount importance.
64
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
To strengthen the JOB it is critical to reinstate its original mandate to review the legality and
propriety of intelligence activities. The board should be given a full staff and should meet regularly.
It must also have a carefully chosen membership, with special emphasis given to those individuals
with enough experience and political clout to avoid resistance and being misled. The original
stipulations that the board would hear anyone on an informal basis and with the promise of anonymity
must be reinstituted as well. Under no circumstances should the board get involved directly with an
agency's operation. The IOB is a tool of the President, and it should report directly and exclusively to
him on a confidential basis.
2) The roles of General Counsel and Inspector General should be strengthened and expanded.
These officials must perform in the capacity of ombudsmen. Intelligence agencies' general
counsel must be made aware of all significant activities to ensure compliance with laws and executive
orders. Additionally, the counsel should be situated within the Langely complex and provisions
should be made for confidentiality for employees who come before him with questions or concerns. \
The counsel and his staff should be involved directly with agency affairs and coordinate at all times
with the agency's Inspector General who should initiate systematic as well as particular
investigations. These investigations must meet with total cooperation. To strengthen the Inspector
General's investigative authority and ability to overcome resistance, he should hold rank equal to that
of the Agency's deputy directors. Inspector Generals should also use their office as a forum where
grievances can be heard. If an activity's legality or propriety is questioned, the Inspector General
should bring the matter before the IOB and possibly the Attorney General.
65
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Congressional Oversight.
Congressional oversight of intelligence is almost definitionally problematic. On the one hand
intelligence is a government program funded by and accountable to Congress, but on the other hand it
can be a tool in the execution of policy, as is the military. The Constitution gives both the executive
and the legislature powers in foreign policy, but provides little guidance on jurisdictional line-drawing
concerning intelligence. Congress controls the activities of our government through its power of the
purse, yet it should never interfere with the President's authority as Commander in Chief in a way
which prevents him from acting in the national interest.
Though Congress has the power to make certain intelligence actions illegal it is important that
Congress be wary of using this power in light of Constitutional concerns. The President is granted
much latitude in issues of foreign policy: he may negotiate treaties without the advice of Congress and
has the power as Commander in Chief to engage the armed forces of the United States as long as he
does not take any action which is, in fact, a declaration of war. A mechanism for Congress to prevent
the President from using the intelligence community in a way it disagrees with is by passing
legislation that makes actions illegal. The problem with this is that any time Congress passes a law to
prevent a specific present action it may be proscribing the action from later use when it could be
highly beneficial. Broad prohibitions of this nature are not only bad from a pragmatic point of view,
but they infringe on the President's Constitutional power as Commander in Chief. Only the President
has the power to act instantly on behalf of the United States in the international realm, so it would be
folly to tie his hands ahead of time. If Congress is to outlaw courses of action its prohibitions should
be quite narrow and specific in its legislation. Though this might give a President the potential to
abuse his power, any damage would be limited and quickly controlled in light of the recommendations
below.
Though we take quite seriously the separation of powers, there can be no meaningful oversight of
intelligence unless Congress is kept well informed. Testimony on the Iran-Contra Affair has revealed
66
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
ways in which the executive was able to avoid reporting information to Congress without flagrantly
violating the law. Under the present system, the intelligence committees authorize very specific line
item expenditures for the next fiscal year through the budget process. Any CIA covert activities
which were not authorized in the budget must be authorized by a Presidential finding, and this finding
must be sent to the committees in a "timely fashion." The ambiguities regarding "timely fashion" and
the limitation of the onus to report to actions undertaken only by the CIA have proved problematic in
the past, enabling the actors involved on the executive side to define the word timely to meet their own
needs, and to use agencies other than the CIA to circumvent the responsibility of issuing a finding.
A second problem that has been manifested in recent years has been the tendency for executive
actors to use a loophole in the arms export act that enables the CIA to ship arms in large quantities to
destinations of their choice without the approval of any other agencies of government. We do not
believe that this is a necessary or warranted privilege and have devised recommendations that would
ensure that the law coincides more closely with the intent of its drafters. Covert shipments through
the CIA presently allow avoidance of the much tougher requirements of the Arms Export Act.
According to current law, the President must only report arms shipments which contain a single item
costing over $1,000,000. A President could thus ship an infinite amount of small arms anywhere he
pleases. Recommendation number 2, below, addresses this concern.
One of the powers that Congress presently enjoys is the authority to release information to the
public. It is important to the principles of a democracy that Congress, made up of the people's
representatives, have the power to determine when releasing classified information is in the best
interest of the people. However, disclosure to the public should be discouraged and should not take
place in a way that jeopardizes American security or American lives; a real danger in situations when
Americans are engaged in clandestine operation abroad. Especially in light of the fact that Congress
will know about a given action within two days of its inception at the latest, we believe that some new
expression of the concern with which we view disclosure is mandated. Recommendation 3 follows.
In light of the above concerns, we suggest the following:
67
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
1. In the area of notification, the President should be required to issue findings for all unbudgeted
covert activities regardless of the agency carrying them out. Additionally, in lieu of the present
standard of "timely fashion" we recommend that the findings be filed with the appropriate members
of congress within 48 hours of the actions they concern. We also recommend that actions not be
authorized retroactively, and that findings be in writing. If there is not enough time to put a finding
in writing before acting, therefore, the President has 48 hours to put the finding in writing and
submit it to the appropriate members of Congress.
2 Regarding the shipment of armaments, the conference recommends that arms shipments be
treated in the same manner as all other covert actions. The $1,000,000 single item floor triggering
notification should be replaced by a requirement that all previously unbudgeted covert arms
shipments trigger Presidential findings. These findings should also be sent to Congress in writing
within 48 hours of the shipment.
3. Because the two recommendations above and their provisions for immediate disclosure to
legislators will give the Congress more control over the President's use of the intelligence
community, we recommend that if a house of Congress votes to release information to the public, it
wait 30 days before it actually releases the information. This should give the President time to
dismantle any operation to be revealed and should do something to keep that balance between the two
branches of government operating in the best interests of all.
68
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Public Opinion and the Intelligence Community.
American public opinion serves an important, if ill-defined, role in the oversight of intelligence.
Public opinion affects intelligence in several fundamental ways: through recruitment, interaction with
academic resources, morale within the intelligence community, external cooperation and support for
intelligence activities, in using constituent and media pressures, and through influence over
Congressional oversight mechanisms.
It is ironic that despite this panoply of effects, the majority of the American people are unaware of
either the necessity of intelligence functions to the maintenance of national security or the influence
that they wield over those functions. In the face of an absence of official educational information,
public opinion towards intelligence is generally gathered from the mechanisms of popular culture --
televisions, films, and novels about intelligence activities -- and from the mass media. The mass
media since the mid-1960s, however, has presented an imbalanced view of intelligence, tending to
focus on covert action rather than collection and analysis and failures rather than successes. Public
opinion, therefore, has been inadequately informed and, as a result, has tended to equate intelligence
organizations with covert action policy. Polling data reflect a generally negative and cynical public
attitude towards the CIA and its operations.
This negative public attitude has damaged intelligence in several fundamental areas and, by
denying the intelligence community important resources, has impeded the effective operation of
intelligence. Its deleterious effect on the recruitment of intelligence officers has already been
discussed. Negative public opinion has damaged employee morale within the agency, as well. By
attaching a stigma to all relations with intelligence operations, negative public opinion has also
crippled CIA interaction with academic resources and other external organizations, such as American
businesses, which could provide valuable assistance to intelligence operations. Furthermore,
although difficult to calculate precisely, it is generally acknowledged that negative public and media
pressures on Congress have resulted in increased controls on intelligence in the past; if such controls
are imposed unnecessarily in response to uninformed public opinion, intelligence may suffer further.
69
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
There is a need to educate American public oversight of intelligence to make sure that such
oversight acts to influence intelligence operations in a responsible and positive fashion. An
appreciation of the functions and necessity of an intelligence service will serve as a "safety net" to
preclude the recurrent possibility that a temporary lapse in public support for government policy will
threaten national security by hindering the effective operation of intelligence functions.
Though we recognize the need for secrecy, our research has led us to believe that the present
overemphasis on keeping the public uninformed does not strike the appropriate balance, nor is it in the
best interests of the agencies involved or the American public. For example, a study by the National
Security Agency's Information Security Office concluded that in 1980 eight hundred thousand
documents were classified "unnecessarily". We trust that once the significance of the public and the
media are recognized, a more appropriate balance will be found.
Public opinion is a powerful mechanism that wields considerable power over the effective
operation of intelligence operations, and public opinion towards intelligence must be "decoupled"
from public opinion about covert action and foreign policy. Education about the separate and distinct
functions of intelligence information collection and analysis is also warranted. The current
mechanisms for public education about intelligence are inadequate; the Office of Public Affairs at the
CIA presently engages in no public relations activities. To rectify this inadequacy, we suggest a
-public education and information dissemination program, to be implemented on three fronts:
1)Through a CIA-operated "classic" public relations campaign similar to that used by the US
Navy. Designed to educate the public of all levels about intelligence, such a campaign would expand
the Office of Public Affairs and would employ means such as public lectures, conferences, and
interviews; outreach to various community groups through information squads; programs at Langley
to educate elites (like those initiated by Admiral Turner during the late 1970s); education in
elementary schools; media programs; and increase in the number of "successes" broadcast by the
70
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
agency; and a movement to declassify "safe" material in an attempt to educate the public.
2) Through an increased policy of openness and friendlier relations with the media, in which a
relaxation of antagonism might lead, as it has in the past, to a more mutually beneficial relationship,
where the media, involved in a channel of communication, might educate the public in a more
balanced manner about intelligence issues.
3) Similarly, through open, friendly relations with its Congressional oversight committees, the
CIA and other intelligence agencies may establish a working situation where those committees, by
publishing positive reports about intelligence operations, would act as alternate, highly credible,
informants of American public opinion.
71
im Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Summary of Recommendations.
I) Further enhancement of informal modes of communication, rather than a more formal restructuring of
the hierarchy, will most productively reduce impediments to information flow within the intelligence
community.
2)We recommend that intelligence agencies offer extended exposure to one region or issue during the
early stages of their careers and only then rotate their assignments.
3) In order to reduce the problems associated with information overload, we recommend filtering the
flow of data up through the intelligence bureaucracies.
4) While it is important to educate policymakers about the possible bias in the work of the various
intelligence agencies, no attempt should be made to correct the institutional biases of the three main
analytical agencies.
5)We recommend preserving the dual roles of the DCI.
6) The proper degree of authority is presently invested in the DCI.
7) The DCI must function as an unbiased reporter of intelligence analysis, not as a policy advocate.
8) The post of DCI should not be elevated to cabinet level.
9) The DCI mast be allowed adequate flexibility to shape the intelligence bureaucracy as he sees fit..
10) Historical observations of past DCIs should serve as guides to future appointments.
11)The Military Intelligence Community should be significantly restructured.
12) DIA should focus on improving the quality of its personnel.
13 )The Military Intelligence Service should implement new programs for continuing education.
14) INR should preserve and enhance close contact between desk officers and country analysts.
15) INR should continue to include Foreign Service Officers among its analysts.
16) INR should place more emphasis on pro-active and long-term analysis.
17) INR should be classified as a National Security Agency.
18)We believe that a counter-deception unit might effectively combat intelligence deception.
19) The training of analysts should emphasize an awareness of the problems of bias..
72
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
-ow
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
20) The evaluation process should include a detailed review of the though process of the analyst
21) Important analytical work should be evaluated by a committee composed of both former and active
members of the intelligence community and outside experts.
22) A significant increase in the number of CIA recruiting offices and recruiters is warranted.
23) The testing procedures need to be more comprehensive, ideally involving a writing sample and
an oral examination, as well as an increased focus on foreign language skills..
24) The CIA should increase the amount of training given to analysts.
25) The CIA should provide more supergrade positions for analysts and thereby increase the number of
high level people doing analytical work.
26) Both communities should draw on the high degree of informal contact that already exists to further
increase producerlconsurner contact.
27) Action should be taken to increase contact between middle level analysts and working level
intelligence consumers.
28) Contact between middle level analysts and high level policymakers should be increased.
29) There should be an increased emphasis on long term intelligence reporting.
30) The Central Intelligence Agency should design its covert operations with the assumption that they
will be made public.
31)We recommend that accountability be increased, specifically at the higher levels.
32). More attention should be paid to the hiring and training of Clandestine Operations personnel.
33). We urge that the use of third parties in U.S. operations be strongly discouraged.
34). We recommend that clandestine operations remain vested only in the Central Intelligence Agency.
35) Policymakers should be extremely wary of employing covert action, for the future ramifications
of large-scale military secret operations can and often are extremely serious. Covert action is not
practical; its short-term effects can vary, as demonstrated by the operations in Cuba and Guatemala,
but its overall ramifications are usually quite damaging to the interests of the United States.
36). The President's Foreign Intelligence Advisory Board (PFIAB) and the Intelligence Advisory Board
(JOB) should be strengthened.
37) The roles of General Counsel and Inspector General should be strengthened and expanded.
73
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
38). In the area of notification, the President should be required to issue findings for all unbudgeted
covert activities regardless of the agency carrying them out. Additionally, in lieu of the present
standard of "timely fashion" we recommend that the findings be filed with the appropriate members of
congress within 48 hours of the actions they concern. We also recommend that actions not be
authorized retroactively, and that findings be in writing. If there is not enough time to put a finding in
writing before acting, therefore, the President has 48 hours to put the finding in writing and submit it
to the appropriate members of Congress.
39) Regarding the shipment of armaments, the conference recommends that arms shipments be treated in
the same manner as all other covert actions. The $1,000,000 single item floor triggering notification
should be replaced by a requirement that all previously unbudgeted covert arms shipments trigger
Presidential findings. These findings should also be sent to Congress in writing within 48 hours of the
shipment.
40). Because the two recommendations above and their provisions for immediate disclosure to
legislators will give the Congress more control over the intelligence finger of the President's foreign
policy hand, we recommend that if a house of Congress votes to release information to the public, it
wait 30 days before it actually releases the information. This should give the President time to
dismantle any operation to be revealed and should do something to keep that balance between the two
branches of government operating in the best interests of all.
41) Through a CIA-operated "classic" public relations campaign similar to that used by the US Navy.
Designed to educate the public of all levels about intelligence, such a campaign would expand the
Office of Public Affairs and would employ means such as public lectures, conferences, and interviews;
outreach to various community groups through information squads; programs at Langley to educate
elites (like those initiated by Admiral Turner during the late 1970s); education in elementary schools;
media programs; and increase in the number of "successes" broadcast by the agency; and a movement
to declassify "safe" material in an attempt to educate the public.
42)Through an increased policy of openness and friendlier relations with the media, in which a
relaxation of antagonism might lead, as it has in the past, to a more mutually beneficial relationship,
where the media, involved in a channel of communication, might educate the public in a more balanced
manner about intelligence issues.
74
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0
43) Similarly, through open, friendly relations with its Congressional oversight committees, the CIA and
other intelligence agencies may establish a working situation where those committees, by publishing
positive reports about intelligence operations, would act as alternate, highly credible, informants of
American public opinion.
75
Ns Declassified in Part - Sanitized Copy Approved for Release 2013/01/14: CIA-RDP90G01353R001700010025-0