LETTER TO WILLIAM WEBSTER FROM GREGORY F. TREVERTON

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP89G00643R001000020004-0
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 8
• Document Creation Date: December 23, 2016
• Document Release Date: November 10, 2011
• Sequence Number: 4
• Publication Date: November 13, 1987
• Content Type: LETTER

Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89G00643R001000020004-0
HARVARD UNIVERSITY
J01-IN F. KENNEDY SCHOOL OF GOVERNMENT
79 JOHN F. KENNEDY STREET
CAMBRIDGE, MASSACHUSETTS 02138
Gregory F. Trever[un
Lecturer in I'uhlic I ulicv
Senior Research Associate,
(',enter for Science and International Affairs
STAT
Judge William Webster
Director of Central Intelligence
Central Intelligence Agency
Washington, D.C. 20505
I know that you have been dealing with your. colleagues, especially
about our Intelligence Assessment and
Policy Program, but I wanted to close the loop from Harvard's end. Let
me repeat how much we look forward to the program -- by now, I imagine
you have a read-out from the first two executive seminars. we have held
here in Cambridge. We also look forward to your participation in the
kick-off dinner for the Program's Steering Committee, in Washington on
December 14th. I know I speak for my colleagues as well, especially
Graham Allison, Ernest May and Richard Neustadt.
For your information, I enclose the list of those on the Steering
Committee, Graham's letter of invitation for the dinner, and a brief
description of the Program that was sent with the letter.
We look forward to seeing you on the 14th. With all best wishes,
Er.et,'+.li^ I F i'Sirjl
87-3662x
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89G00643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
HARVARD UNIVERSITY
JOHN F. KENNEDY SCHOOL OF GOVERNMENT
CAMBRIDGE, MASSACHUSETTS 02138
STEERING COMMITTEE
INTELLIGENCE ASSESSMENT AND POLICY PROJECT
Co-Chairmen:
Dean Graham Allison, Kennedy School of Government, Harvard University
Judge William Webster, Director of Central Intelligence
Invitees for Membership:
Morton I. Abramowitz, Assistant Secretary of State, Intelligence and
Research Bureau
Michael Armacost, Undersecretary of State for Political Affairs
Senator Bill Bradley, Senate Select Committee on Intelligence
Dr. Zbigniew Brzezinski, Counselor, Center for Strategic and International
Studies
Frank Carlucci, Assistant to the President for National Security Affairs
Congressman Richard Cheney, House Select Committee on Intelligence
Robert Gates, Deputy Director of Central Intelligence
Lt. Gen. Edward Heinz, Director of Intelligence Community Staff
William Hyland, Editor, Foreign Affairs
Admiral Bobby Inman, Westmark Systems, Inc.
Professor Robert Jervis, Columbia University
General David Jones, former Chairman of the Joint Chiefs of Staff
Richard Kerr, Deputy Director for Intelligence, Central Intelligence
Agency
Andrew Marshall, Director of Net Assessment, Office of the Secretary of
Defense
Congressman Dave McCurdy, House Select Committee on Intelligence
Senator Sam Nunn, Senate Select Committee on Intelligence
Lt. Gen. William Odom, Director of the National Security Agency
Lt. Gen. Leonard Perroots, Director of the Defense Intelligence Agency
James Schlesinger, Counsel, Center for Strategic and International Studies
General Brent Scowcroft, former Assistant to the President for National
Security Affairs
George Shultz, Secretary of State
Caspar Weinberger, Secretary of Defense
James Woolsey, Shea & Gardner
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
HARVARD UNIVERSITY
JOHN F. KENNEDY SCHOOL OF GOVERNMENT
GRAHAM ALLISON
DEAN
nm
add
79 JOHN F. KENNEDY STREET
CAMBRIDGE. MASSACHUSETTS 02138
I am writing to invite you to become a member of the Steering
Committee of a-Project on Intelligence Assessment and Policy, being
conducted by the Kennedy School of Government. The goal of the project,
funded by the Central Intelligence Agency, is to improve the interaction
between intelligence and policy in assessing foreign governments and
making foreign policy. To that end, the project comprises three sets of
activities: developing case studies on intelligence assessment and
decisionmaking; undertaking a series of one-week training programs at the
Kennedy School for intelligence analysts; and organizing a series of study
group meetings on specific issues of intelligence assessment and policy.
The Steering Committee consists of high-level current and former
government officials from both the intelligence and policy communities,
as well as experts from the academic community. We anticipate that this
Committee will meet twice a year to discuss general issues in
intelligence/policy relations and to provide advice and recommendations
for the development of project activities. A list of those we are
inviting to become members is attached, along with a project description.
We would be honored to have you as a member of this Steering Committee
and to join us for its first dinner-meeting, December 14, 1987, from 7:00
to 10:00 p.m., at Bacon House, 1801 F Street, in Washington. Judge
Webster and I will speak briefly about the project. Professor Ernest May
of the Kennedy School will then provide a further description of the
project and moderate the discussion.
One of my colleagues directing the project will call you in the next
few days to answer any questions you may have. We look forward to what we
hope will be your favorable response and to your participation in this
important project.
Sincerely,
Graham T. Allison
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
HARVARD UNIVERSITY
JOHN F. KENNEDY SCHOOL OF GOVERNMENT
CAMBRIDGE, MASSACHUSETTS 02138
INTELLIGENCE ASSESSMENT AND POLICY PROJECT
Project Structure and Planning _
The Intelligence Assessment and Policy Project aims to understand better
how assessment relates to policymaking and, especially, how policymakers
do--and might--use assessments. The three-year program of research and
curriculum development is designed to draw on the expertise of both academics
and practitioners. By examining and drawing lessons from a series of
historical and contemporary cases, the project will develop teaching
materials and methods for university courses and training programs for
analysts, and produce a variety of publications.
II. Analytic Categories for Assessments
An idealized process by which assessments assist policymaking might be
the following. First, a situation exists in which policymakers want to
influence or make a decision regarding a foreign government. They need
appraisals of the foreign government in order to understand its make-up,
resources, objectives, governmental processes, etc. In the course of
defining the problem in hand, U.S. objectives, and U.S. options, the
policymakers need estimates of the possible effects of U.S. actions on the
foreign government and hence of new conditions that the United States may
face at some later point. We think of "assessment" as analysis of U.S.
problems and policy choices informed by analysis of how other governments may
perceive their problems and choices. It is hence inherently an activity
requiring cooperation and interaction between decision-makers and experts.
In hope of learning how this interactive relationship might be made more
efficient and effective, we propose to analyze and to invite informal debate
about a variety of cases in which assessment played, or should have played,
an important role in policy decisions. We have established an initial
framework for these analyses by focusing on six abstract "difficulties" an
analyst may face that must be overcome if his/her goal is to be attained.
These difficulties fall into three general categories, defined by the reason
for the difficulty.
A. NATURE OF THE ASSESSMENT
1. The unwelcome news difficulty. In-'this situation, the analyst's
estimate will be bad news to the policymaker, e.g., it provides
intelligence that undermines or calls into question the
policymaker's current policy or inclinations. The difficulty the
analyst faces is that his/her estimate will be, ignored.
2. The too-welcome news difficulty. In this situation, the analyst
knows that his/her estimate will be good news to the policymaker.
However, the difficulty the analyst faces is that the estimate will
.._,co;.e that it will' be used L.. -Ustiiy Eiote comprehensive or
extreme policy actions than can be appropriately supported.
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
B. ABSENCE OF CRUCIAL ELEMENTS
3. Missing market difficulty. In this situation, the analyst has
created what he/she views as an important estimate, but discovers
that there is no apparent market for the news. The difficulty here
is that the estimate will not be utilized, unless the analyst finds
or creates the market.
4. Missing or inadequate product difficulty. In this situation, the
policymaker has specifically requested an estimate or the policy
process itself is clearly in need of it, but the intelligence an
analyst needs is missing or inadequate. The difficulty here is that
the intelligence community is unable to fill its assigned role of
contributing to policymaking. Policy may thus be made in ignorance
of vital information, and the reputation of the intelligence
community itself could be harmed.
C. ORGANIZATIONAL SETTING
5. "Intragovernmental" difficulty. In this situation, the analyst
provides an estimate in which he/she has confidence, but the
different policymakers who receive the estimate have different
policy preferences, and the estimate becomes a pawn in their
internal games.. Here, the estimate may eventually be well-utilized,
but the difficulty is that it may be misused, ignored, or used
selectively or only after delay.
6. "Transition" difficulty. In this situation, a newly appointed
official or a new presidential administration faces a policy
decision, but the newcomer or newcomers are unfamiliar with the
structures and offices of the government and not only may not know
which questions to ask but also may not know how to go about finding
the necessary answers. The difficulty for the analyst is, again,
that policies will be made in absence of adequate assessment, and
while it may not do harm to the intelligence community, it will
deprive it of early opportunities to make a positive contribution.
The cases to be examined will be chosen on the basis of potential for
helping to answer the question, "What makes for assessments that are better
than usual?" Partly for sound logical reasons but also because much prior
study has focused on mishaps, other cases will contribute by answering, "What
made for assessments that were worse than usual?" The following list of
questions suggests some of the points to be addressed in each case: Did
education, training, and/or career pattern seem to make any difference? What
did individuals on the policy side stand to gain or lose from acquiring
working expertise about a foreign government? Did the nature of organization
or process make a difference? How were analyses of foreign governments
affected by the make-up of the analyzing agencies and the mix of types of
intelligence (human, signal, etc.) or types of expertise (regional,
functional, etc.)? Were overall assessments better to the extent that
intelligence analysis was kept separate from decision-making or to the extent
that the two meshed? Who asked for,(or got) what kind of information, when,
and to what result? Did experts and policymakers have different judgments
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
about how to communicate with or influence the foreign government? If so,
why? To the extent one can judge, who was more nearly right? What
propositions are most nearly validated by experience? How and in what
circumstances have experts induced alteration in the preconceptions of
policymakers? How and in what circumstances have policymakers induced
alteration in the preconceptions of experts?
The cases will be crosscut in four categories defining the relationship
between the assessing government and the foreign government: familiar
friends, unfamiliar friends, familiar foes, and unfamiliar foes. Possible
cases are listed under these categories below, although the process of
examining them will help to determine if their initial placement is correct
or if they, in fact, fit into more than one category. For example, a
situation that a policymaker might characterize as "Unwelcome News" may look
to an analyst like news for which he faces a "Missing Market." In addition,
as cases are studied, the categories may well need to be refined.
DIFFICULTIES
Familiar
Friends
Unfamiliar
Friends
Familiar
Foes
Unfamiliar
Foes
Unwelcome
News
Suez '56
Marcos
Shah
France '40
Iran/contra
Too-welcome
News
France '36
Hitler
tractable
Arms control
treaty violations
Missing
Market
Skybolt
Vansittart
(UK 30s)
PRC '50
Missing or
Inadequate
Product
INF '78/9
Yom Kippur War
Shah
Intragov.
Politics
Six-Day War
(US)
Lebanon
'82/3
Six-Day War
(Israel)
Transition
Truman & Japan
SALT '77
Bay of Pigs
All the historical situations above are possible candidates for case
studies. We will also solicit ideas for cases from analysts in the
project's training programs and Council participants. Re-expect to prepare
twelve to eighteen cases; for the immediate future, we are focusing on five.
The first is a comparison and contrast of the contribution of
assessments to policymaking during the 1967 and 1973 Arab-Israeli wars.
Although the two wars were similar, the first was an intelligence and
policymaking success for the U.S. and Israel and the second was a failure,
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89G00643R001000020004-0
so we expect to be able to draw some instructive contrasts. The case will
examine both U.S. and Israeli assessments off. the situation.
The second will focus on the fall of the Shah in Iran and its
aftermath. Drawing on the excellent literature on the subject and the
reflections of participants, the case will make explicit the lessons from
this intelligence failure. The case may also be broadened to consider more
generally the question of why the U.S. has had some successes but mostly
failures in its history of assessing Iran.
The third will examine the performance of the intelligence community in
the Iran-contra case. This case will focus especially on the
interaction of the estimating process with the politics of the situation.
The purpose of studying this case will, of course, be pedagogic rather than
investigative.
The fourth case will consider the Vietnam war--an intelligence success
but a policymaking failure. The case will focus on high-level assessment in
Washington, especially the evidence and procedures that produced accurate or
inaccurate assessments of Vietnam and of the United States.
The fifth case concerns the French in 1940, when extraordinarily
accurate estimates of German military plans seem to have been known to
analysts but totally ignored or rejected by decision-makers.
The project will be shaped in part by a series of consultations with
current and past senior officials involved in intelligence assessment,
policymaking, or congressional oversight. This group, known as "The
Council," will consider plans for the project, review its research and case
products, and act as both resource and sources of guidance to the effort.
The Council will include a select steering group of about twenty members
co-chaired by the Dean of Harvard's Kennedy School of Government and the
DCI. This steering group will meet twice yearly to provide overall guidance
and review of the project. Steering group membership will include senior
intelligence community leaders, current or past senior officials with policy
roles, and members of Congress involved in oversight of intelligence
activities.
In addition, a working-level group will meet more frequently (four to
six times each year) to consider particular research products, cases, and
other aspects of the project. The working-level group will be somewhat
larger (25-40) and will include a core group of regular members. Other
individuals with specialized knowledge of the particular topics under
discussion will be invited to each meeting.
The format of the meetings may vary, but will generally consist of a
late afternoon discussion, followed by a working dinner. Most meetings will
be held in Washington.
The first council meeting--which will be for steering group members and
Kennedy School project directors--is scheduled for Monday, December 14,
1987, in Washington. This meeting will include presentations by the Harvard
participants of the outline of the overall project and an opportunity to
discuss its content with the steering group. Dean Graham Allison and Judge
William Webster will co-chair this initial meeting.
Coordination and arrangements for all of these meetings are the
responsibility of the School's project staff.
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89G00643R001000020004-0
 Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0
V. Training Programs
The next one-week training program for mid-level intelligence analysts
will be held at the Kennedy School the first week of November. Project
members have taken careful note of the participants' evaluations of the
first, experimental program, and are endeavoring to take account of their
primary concern that the cases and discussion more explicitly address the
role of intelligence assessment in policymaking. The discussions will also
attempt to draw more directly on the experience of the analysts
participating. Some existing'cases will be modified for this program, and
new cases are being prepared.
:The project is directed by Kennedy School faculty members Ernest May,
Richard Neustadt, Gregory Treverton, and Peter Zimmerman. Administrative
assistance is provided by Nancy Huntington; research coordinator is Lynn
Whittaker.
Declassified in Part - Sanitized Copy Approved for Release 2011/11/10: CIA-RDP89GO0643R001000020004-0