PFIAB STUDY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP91M00696R000700010005-4
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 3
• Document Creation Date: December 15, 2016
• Document Release Date: March 1, 2004
• Sequence Number: 5
• Publication Date: May 5, 1976
• Content Type: MF

~I..rJ~14JJ L A
Approved For Raioeise 2004/05/13: CIA-RDP91 M00696R9700010005-4
S May 1976
MEMORANDUM FOR: Mr. Knoche
Mr. Proctor
SUBJECT PFIAB Study
1. Attached is a copy of the paper submitted to the
PFIAB by the three-man subcommittee -- chaired by Bob
Galvin -- which the Board appointed last December to
scrutinize the Community's performance in strategic
estimates. As you know, the other two members of the sub-
committee are Drs. John Foster and Edward Teller.
2. In my opinion, it is essential that we not be
unduly defensive in reacting to this document or in any
way convey the impression that we are loathe to consider
innovative procedures which might improve the quality of
our strategic assessments. On the other hand, some of_
the concrete proposals advanced by the Board would be
extremely difficult to accommodate without prostituting-
the whole intelligence process. This applies particularly
to the proposal (about which the Board feels very strongly)
for a "competitive analysis group" which would be tasked
with preparing -- on certain selected issues -- what
would in effect be an alternative estimate to 11-3/8-76.
3. The basic problem is the fact that the Board's
concrete recommendations derive from a perception of
what intelligence is and ought to be which is quite
different than ours -- and in this case, I am personally
convinced we are right and the Board, wrong. The game
(unintentionally) is given away in three sentences in the
NIE Evaluation Committee's paper:
Approved For Release 2004/05/1
SECRE'
25X1
25X1
25X1
 25X1
Approved For Release 2004/05/13 : CIA-RDP91 MO 6R000700010005-4
-- "...whether or not a particular technical
judgment in the NIE is correct or incorrect
is less significant than whether the
document illuminates for a busy decision
maker the range of threat possibilities and
their implications relative to his special
responsibilities." (Paragraph 6, pp 2-3).
-- [One of the four purposes an NIE serves is-
to] "Support Congressional authorization and
appropriation proceedings." (Paragraph 17b,
p. 6. A similar thought is echoed in paragraph
29 on page 9: "...during Congressional
hearings, the NIE may present serious
problems to Defense officials whose programs
are based on different threat appraisals.")
-- [NIEs] should be measured by whether they
stimulate policy makers to face up to hard
decisions in sufficient time to make a
difference and by the thoroughness with
which threats, uncertainties and alternatives
have been illuminated." (Paragraph 34, p. 11).
4. What the Board wants--is a -national estimate which
will-.set forth all the things --,especially the unpleasant
.things ---.which the Soviets could or-might do, without
any estimative. judgments about-the relative probability
of the Soviets:achieving these various goals or pursuing
these alternative lines of behavior. The real reason
(I think) why some members of the Board are pushing for
"the competitive estimate" by a group composed of at
least some persons outside the Intelligence Community.
is that they want to be sure that the total package
includes all the worst case possibilities that can be
thought of. Under the approach the Board is recommending,
the President and his senior policy advisors will simply
have this range of possibilities laid before them, hence,
powerful arguments could be advanced that the only re-
sponsible course to follow to protect the nation's
interests would be to hedge against the worst case
threats, and NIEs developed through the recommended
procedure would serve as ammunition supporting such a
pitch. If our nation's resources were infinite, this
Approved For Release 2004705/13 : CIA-RDP91 M00696R000700010005-4
SECRET4
 might be an intellectually defensible thesis. They are
not, however and, hence, it isn't. This procedure would
leave the decision maker at the mercy of technical shamans
with no basis for ascertaining which of these shamans'
analyses or predictions were more credible than their
competitors'.
S. Per our discussion at the 5 May Executive Committee
meeting, I urge that copies of the attached paper not --
repeat not -- be circulated. Its language, in many
places, is outrageous and (with reason) would be taken
as deeply offensive by many hardworking professionals who
are fully conscious of their own fallibility but have
devoted their careers to providing our government with the
most objective and balanced assessments attainable by
mortals operating with less than total information. Bob
Calvin, an eminently decent person trying to do a very
serious job, is most anxious to avoid initiating a paper
war for reasons I consider obvious, commendable and entirely
persuasive. I plan to talk quietly with him in order to
try to separate the concrete recommendations -- several of
which are well worth trying -- from the philosophy behind
some of them which we cannot endorse. I am reasonably
confident that through quiet dialogue we can develop a
set of experiments which may actually prove helpful, will
put the DCI in the posture of being responsive to the board
and which, at the same time, will not sacrifice our
principles or. things all of us believe in very deeply.
Approved For VA0ease 2004/05/13 : CIA-RDP91 M0069e e00700010005-4
25X1
25X1
George A. Carver, Jr.
Deputy for National Intelligence Officers
O/D/DCI/NIO:GACarver/kes
Distribution
Orig.- Mr. Knoche
1 - Mr. Proctor-
1 - DCI (copy of Galvin note to Carver only)
1 - D/NIO Chrono w/att
1 - PFIAB file w/att
1 -
Approved For Release 2004/05/13 3C - 696R000700010005-4
SECRET/
25X1