KIQS AND KEPS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP91M00696R000500080007-7
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 6
• Document Creation Date: December 15, 2016
• Document Release Date: February 17, 2004
• Sequence Number: 7
• Publication Date: February 7, 1975
• Content Type: MF

Approved For Release 2004/05/13 : CIA-RDP91 M00696R000500080007-7
7 February 1975
MEMORANDUM FOR:
SUBJECT KIQs and KEPs
1. As you are all well aware, we are still
wrestling with the problem of devising an evaluation
report which the NIOs can prepare after the end of a
fiscal year, reviewing and auditing the Community's
performance on each Key Intelligence Question. The
problem is to devise an approach, hence a form, which
will be manageable (in terms of the work required to
pull it together) and will still give the DCI the kind
of information and evaluation he needs to have a clear
idea of how the Community has done, how efficiently its
efforts have been orchestrated, how well its resources
have been applied, and what changes (if any) in direction
or resource allocation he ought to initiate, encourage
or -- where he can do so -- direct in order to improve
the Community's performance. As you all are also well
aware, there is a dichotomy between the approach
instinctively preferred by this office and that advo-
cated by some components of the IC Staff, notably
2. We have finally found a format embodying a
general approach the Director considers promising:
namely, a fourth try at a KEP on KIQ 17-B of the FY 1974
Pilot Program prepared by
a copy of.which is attached. is Model, which the
has endorsed, is nonetheless perhaps atypical in one
critical respect: it reviews the Community's performance
on a KIQ on which the Community, as a whole, performed
rather well and on which there are no serious short-falls
either in collection or in production which require
25X1
Approved For Release 200 IA-RDP91 M00696R000500080007-7
. ff ftf
 SECRT
Approved For R ease 2004/05/13 : CIA-RDP91 M00696'00500080007-7
attention and corrective action. Furthermore, the
I)CI is still not quite sure that this line of
approach gives him the resource information --
hence, handle -- he wants and feels he needs. Also,
this being the case, he remains susceptible to
Is contention that the full-bore systems
analysis approach which has advocated
from the outset is what e DCI needs and, further, to
the argument (which has also consistently
advanced) that this approach can be followed without
generating an undue amount of work or putting undue
strains on the Community or its member components. At
a minimum, the DCI is intrigued with
matrix, the FY 75 version of which is
your inspection.
3. It has not been easy to move toward a
realistic resolution that meets the DCI's needs but
at the same time takes cognizance of the realities
of the amount of work involved in assembling various
types of data and the actual utility of this data,
once assembled. In part, progress in this field has
been complicated by penchant for seeing
the DCI` privately and making a presentation which has
not been well staffed even within the IC mechanism, let
alone with this office.
4. One reason why I kept and
logging at successive rewrites of a KEP for
that I thought I had an understanding with
the DCI that we would try to develop a KEP model he liked
for one FY 74 KIQ, after which we would prepare three
more illustrative examples (on three additional KIQs in
different areas). He would then have a quartet of
examples employing the verbal "fuzzy" approach which he
could compare with the type of KEP developed through
techniques. With these two sets of
parallel examples simultaneously before him, he could
then decide which best met his needs and, further, whether
the latter was worth the added marginal cost of effort
entailed in its production.
S. Last Saturday, the above plan got partially
derailed, again by -- who somehow
convinced the DCI that this would not be a valid comparison
Approved For Release 2004/I -RDP91 M00696R000500080007-7
25X1
25X1
25X1
 Approved For release 2004/05/13 : CIA-RDP91 M00696VO0500080007-7
or a fair one with respect to the KEPs developed through
the systems analysis type of statistical approach.
The thesis here (which I do not regard as valid but which
the DCI accepted) was that the preparers of the "fuzzy"
KEPs had the advantage of the information assembled
through the full-bore process, hence a comparison of
the two results was not a fair test. Instead,, Fuzzy KEPs
should be prepared on FY 74 KIQs that were not part of
the Pilot Program so that the NIOs preparing them would
not have this unfair advantage. I argued -- with, I
fear, minimal impact -- that this approach also had some
built-in problems even more serious in nature. Though
a full set of 31 KIQs had been developed for FY 74 to
show everyone what a full set of KIQs might look like,
in point of fact, only nine were actually worked on at
all, these being the nine on which an IC Staff exercise
had been done. Fuzzy KEPs would be almost impossible to
prepare on any of the remaining 22 since there had been
no interagency discussion of them in detail and any such
model KEP would have to be sucked by an NIO out of his
thumb.
6. General Wilson and I discussed this whole
problem on 5 February. We agreed that it was essential
that his colleagues and mine try to work together to
develop a common proposal to lay before the DCI: a
proposal which should be a single package even if it
reflected certain unresolvable differences of opinion or
recommended approach between our two offices. To that
end, we agreed to caucus at 1500 on Monday, 10 February,
on neutral ground (the DCI Conference Room).
7. The purpose of this memorandum is to recount
the background cited above and to invite all five of
this memorandum's addressees to attend that session with
me. It would be articularly helpful if
I Icould block out a KEP for one FY 74
Pilot Program KIQ in each of their respective areas (e.g.,
1-D, 21 and 15-A) along the lines of
model,which the DCI has endorsed. General Wilson and I
both hope that this kind of joint discussion between
SECRET
25X1
25X1
Approved For Release 2004/ORDP91 M00696R000500080007-7
 SECRET
Approved Forelease 2004/05/13 : CIA-RDP91 M0069W000500080007-7
ourselves can lead to a resolution of this problem and
an end to the confusion that has been engendered
by a series of unilateral approaches to the DCI.
George A. Carver, Jr.
Deputy for National Intelligence Officers
cc: Acting Comptroller
25X1
Approved For Release 2004/6 i-RDP91M00696R000500080007-7
 Approved For Release 2004/05/13 : CIA-RDP91 M00696000500080007-7
23 January 1975
MEMORANDUM FOR: The Director
SUBJECT The Fuzzy KEP
(Round 4)
Per nur 21 January conversation,
has re-worked his model
fuzzy performance report, concentrating
on the second half. Attached is the
result of his labors. I f we are still
wide of your mark, please let me know.
As I explained, once we get one model
set as reflecting what you are looking
for, I will then have three or four
more clone using it as a paradigm. We
will then be in a position to see
whether we have an art form likely to
meet your needs and all other consi-
derations.
25X1
George _. Carver, Jr.
Deputy for National Intelligence Officers
Copy No. 1
Copy
Copy
No.
No.
2 - DDCI
3 - D/DCI
/IC
Copy
No.
4 - ICS/K
EP
Pilot
25X1
Copy
Copy
Copy
No.
No.
No.
S - D/NIO
6 - N [O/S
7 - N[O/R
Ch
SEA
I
rono
Copy
No.
S - ER
25X1
Approved For Releas? efl O4110.: CIA-RDP91 M00696R000500080007-7
 25X1 Approved For Release 2004/05/13 : CIA-RDP91 M00696R000500080007-7
Next 2 Page(s) In Document Exempt
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000500080007-7