MEMO TO (SANITIZED) FROM ROBERT M. GATES

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP88-00428R000100090006-8
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 3
• Document Creation Date: December 20, 2016
• Document Release Date: July 13, 2007
• Sequence Number: 6
• Publication Date: November 10, 1982
• Content Type: MEMO

Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8
DDI 49115-82
10 November 1982
Director of Training and Education
FROM Deputy Director for Intelligence
1. of the NSC S taf f came to see me the other
day and volunteered to speak to Agency training courses. He
gave me the attached outline of remarks fie makes when asked
to give such presentations and it strikes me that this highly
ra matic view is a useful one for at least DDI folks to. hear.
has spoken to Agency training courses on previous ocJcasions.
2.~~ has knocked around senior levels of the government
for a number of years and is a long time friend of the Agency. I
send this along, noting that he has a message that he is willing
to deliver and may be worth. hearing periodically.
STAT
STAT
STAT
STAT
STAT
Rober M. Gates
Attachment:
As Stated
Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8
 Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8
Issues in Policy~Research
1. Full of "apples and oran es" issues, with few common denominators (e.g.,
. economic, mi nary, po itical, social criteria = "linkages" = imponderables).
Problems are usually "one of a kind." Issues show up as dilemmas and
sometimes as paradoxes.
2. Startin points: difficulty determining where we are now (current condi-
tions where are we headed (the trend -- but how many events make a
"trend?"). Stating "the problem" is usually impossible.
3. The destination problem: where do we want to go, or what do we seek to
avoid priorities). Where do you begin? Who is to begin?
4. Routes: how do we get to a desired destination = plans, programs, guesses
at consequences, establishing alternatives, and options. "The alternatives"
are never exhaustive.
5. dosed with: belief s,_ values, latent ideology, and moral views.- What is the
place of ideology in problem solving? Importance of both ideology and
analysis.
6. Boundary problems: where does a problem."end." If it ends, something will
be left out. With no boundaries, a study will never be finished.
7. At what point do facts, opinions, data, the tracing of consequences,
analysis, ratios, etc., hurt policy research (i.e., when should IQ yield to
judgment?). When is a fact a "fact?" Common sense: what you use when
there is nothing better to go on.
8. Who decides? Who should decide? When?
9. Use of experts. The expert's batting average: 50%.
(a) who i_s an expert? (identification) _
(b) what is~his view of the problem? (education)
(c) consultants. can be "objective," but don't. feel "the-system" (educated
incapacity)
10. Operational problems: fatigue, rushing; big backlogs; bureaucratic
irritations; many meetings;. people with information remote from those who
need it. Tendency to seek simple answers to complex questions; to confuse
bureaucratic rituals and procedures with problem-solving. Confusion of
inputs with output:.. The high price paid_for incompetence.
Helpful academic tools: math, statistics, economics, physics, psychology,
t e Eng ish anguage, computers, law, some history, logic, and whatever your
pet subject is. It helps to have "hyphenated-interests."
11. qualities needed:- brain of Newton, patience of Buddha; insight of Freud;
footwork of Sugar Ray Robinson. On personal behavior, it never pays to get
rattled.
12. The hardest maneuver: to reverse yourself- after a mistake.
13. Many satisfactions: rewarding and broadening if you enjoy (or can put up
with) c a enge.
Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8
 Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8
THE PRACTICE OF POLICY RESEARCH -- AN ART FORM
1. Types of research: basic; experimental; applied; futures; and policy.
2. Policy research deals with decisions, not subjects; knowledge is a necessary
but not sufficient conditions. Otfier conditions: timing, persuasion, and
(a) issues, (b) problems, ability to handle "policy."
Types: a. Decisions on specific problems
b. Policy guidance (i.e., how to think about a set of circumstances)
c. Topics that need to be studied -- staying ahead of the problem
_ problem statement
wisdom
3. Data (experience)--information (testing) = knowledge ~ alternatives ~-GOALS
increasing uncertainty decisions
STAT
4. Aimed at decision makers, staffs, line operations, rather than professions.
Po icy research is not supposed to contribute to a "body of knowledge.")
5. Results are time-specific rather than "timeless." Results are not closed
so-lotions, but are resolutions (at best).
b. Main question of policy research: asks "what difference. does the research
make;" not how well something is known (a scientific test).
7. Methods: eclectic, broad, heuristic. There is no single method or even a
set of methods.. No textbook, much less a "back of the book."
Policy research is much less disciplined, unified, or rigorous than conventional
academic work. Accuracy: Is the sign correct? Order of Magnitude.
"Tools for the Practioner": Probability, marginal analysis, opportunity cost,
iscounting o uture ows.
Two Worlds:
World A
Aristotle
Newton
Adam Smith
Freud
Thomas Malthus
Euc1 id
The Club of Rome
Economics Department,
University of Chicago
Management by Objectives
World fi
P lato' s Caves
Kafka
Jung
Heisenberg
Goedel
Lewis Carroll
Mah 1 er
M. C. Escher
Catch 22
Beckett, Pinter, Oostoyevsky
Approved For Release 2007/07/13: CIA-RDP88-004288000100090006-8