LETTER TO (Sanitized) FROM GEORGE A. CARVER, JR.

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80R01720R000900070038-1
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 8
• Document Creation Date: December 16, 2016
• Document Release Date: July 28, 2005
• Sequence Number: 38
• Publication Date: June 5, 1974
• Content Type: LETTER

Approved For Release 2005/08/02: CIA-RDP80R720R00QQ0 70~0~3~~1
5 June 1974
me your draft note. It makes
i
ng
Thanks for show
points that need to be made and I would strongly
encourage you to give it to the Director since it
brings up a number of issues on which he should focus.
I do have a couple of comments, but they should
be read in (and only in) the context of my basic
endorsement and imp support of your essential argument.
OCI is too hierarchical. There are too many super-
visory levels between analysts and publications.
Substantive meaning is frequently changed in, detri-
mental ways without the substantively knowledgeable
authors being consulted or having any voice innethe
matter until the deed has been done. The
heulrd not be
structure does need revision. People
forced to become bureaucrats in order to advancefes5ional
professionally -- particularly, the price of p
advancement should not be foresaking whattu oene does what one
well (e. g., substantive analm)of others`
may do indifferently (e.g., management work), especially when ineffective discharge of the
latter function can hamper effective discharge of
an~I aAnalysts
the former by oneformer
greaterdo
involvementcolleagues.
need a greater sense
voice in the end product. There also needs to be a
better and more genuine
on whose labors
levels of management.
Accepting and endorsing all of the above, however,
the picture is not totally one-sided, and the whole
picture needs to be seen in proper perspective
truly effective correcte~action ris toebeotaken.
Perhaps inadvertantly yon r the fact that what
area of continuing difficulty:
"conscientious analysts ... all know" is' not always
accurate, despite the certitude of their convictions.
Approved For Release 2005/08/02 : CIA-RDP80R01720R000900070038-1
 Approved For Release 2005/08/02 : CIA-RDP80RQ 2OR000900070038-1
do of NIO secretaries
,For exam le, one quarter
lane
permits as T result. To echo your own prose, this
is a petty point but indicative of something that
needs to be considered.
Analysts, particularly in OCI, have a number of
legitimate grievances which need address -- as you
point out most cogently. Nonetheless, it is also the
case that many analysts -- including quite a few-in
OCI -- are extremely thin-skinned and prone to take
umbrage where the equities are not quite as one-sided
as they are wont to believe. Many writers are indeed
touchy about their prose and reluctant to accept
either well warranted editorial suggestions or
necessary decisions, such as those required. by having
to fit a 400-word piece into a 300-word space and,
hence, needing to prune 100 words late at night 15
minutes before a printing deadline. Furthermore,
many analysts are more than a little prone to self-
righteousness and quick to impute sinister motives
to legitimate differences of judgmental opinion.
Also, members of the analytic fraternity do sometimes
forget that we do run a compartmented organization
in which everybody does not, and should not, necessar-
ily know everything. In my own former incarnation
I used to derive more than a little wry amusement
from passionate assertions that the Agency had never
been consulted on various issues, when in, fact the -
qubeen
thehmattersactually
key analysts i the tclockinvolved
estion.
working around
Finally, there is a strain of thinking prevalent
in a number of analysts -- again including many in
OCI -- which I regard as quite unhealthy. Ironically,
but perhaps not entirely coincidentally, this was a
prominent theme in that vicious PLAYBOY denegration
of Bill Colby --? the theme that the Agency is divided
into sheep and goats: the pure analysts, totally
virtuous in their objective rectitude and the fallen
operators, an amoral lot with few redeeming virtues.
(Managers, in this context, tend to be lumped with
operators.) In this catechism, analysts (alone) have
convictions, others (especially operators) have
biases. Analysts have subtle, sensitive perceptions;
Approved For Release 2005/08/02 : CIA-RDP80RO172OR000900070038-1
 Approved For Release 2005/08/02 : CIA-RDP80RO1172OR000900070038-1
operators (by definition) do not. Analysts are able;
operators (generally) are not. When they are, that
is also bad because operators tend to find soapboxes
uncongenial and are generally prone to acknowledge
that their superiors do actually have the right to
make decisions and issue instructions with which
subordinates may not always agree.
In sum, you have a case, it is a good one, and
it ought to be put to the DCI with vigorous, forceful
allmon~one complicating
argumentation.
are not some
howeverr,eand total picture
elements, side of the fence.
George A. Carver, Jr.
Deputy for National Intelligence Officers
0/D/DCI/NIO:GACarverAI
Distri ution
Orig - I Iw/att./
1 - D/NI rono att
1 - NIO/RI w/o att
Approved For Release 2005/08/02 : CIA-RDP80RO172OR000900070038-1
 Approved For Release 2005/08/02 : CIA-RDP80RQ.1,720R000900070038-1
Anal st-editor frictions in connection with the NID
Y
as stemming only from analystsf
should not be laughed off
vanit over having their "'beautiful Prose tampered
petty y
which Lehman appreciates
with. The particular problem,
and is working on, is of meanin being changed at night
when the analyst isn't around to be consulted.
The root problem is a declining sense of analyst
onsibilit.y--which has a lot
personal involvement and rasp in OCI. There
to do with the general decline in morale
has been some of this. When I was new in oCI the
always fellow at the next desk said
very bright (and since departed)
ing thinking P,p robably nothing important
he got up each morn
in m area and if it does, someone else will
will happen my QVerTUled, if
handle it." Analysts there are used to being
at least
onl because the place is so damned hierarchical (
y ublicat~.c~,
four levels of supervisors between analyst and p
coming down the chain of command as often as
with changes .ace-to-face). But being overruled without even a chance
f analysts
to argue back is a new element. At the same time,
have lost the Divisional "Staff Notes" as a means for
Those used to go out to
expressing their own ideas.
pre-
other specialists clearly labeled uncoordinated and pre-
oiinaranalyst opinion. Now that they are put in hard
lin Y
Approved For Release 2005/08/02 : CIA-RDP80RO172OR000900070038-1
 Approved For Release 2005/08/01 . CIA-RDP80R0 0R000900070038-1
cover, as a substitute for the NID to consumers not
receiving the latter, the Notes, too, have to be approved
up and down the hierarchy. Unfortunately, many analysts
see the NIOs as yet another layer of bureaucracy between
themselves and their readers (I know that isn`t what you
or the NIOs intend, or how I think that system's developing;
but that has yet to be fully demonstrated to. OCI's analysts).
This is not a problem that ever can-be solved.. Large
organizations have to be somewhat hierarchical. (Which is
one reason I feel strongly that OPR,.or whatever group. you
choose to be the "cutting edge of analysis," must be kept
small.) But there are some changes which'could alleviate
the problem.
1. Brownman's new parking-permit system to encourage
car pooling. Only clock-watchers can car pool.. -Conscientious
analysts who put in a 13-hour day to see an important NID
piece clear through the editorial process id then hike to
the West Parking Lot all know that every single NIO secretary
has lane parking (no nonsense about car pooling therer).
I know this is a petty point; but it's indicative to analysts
of how they rate in your scheme of values.
2. Promotions. Grades are higher, time-in-grade
considerably shorter, for people in the superstructure (NIO
Approved For Release 2005/08/02 : CIA-RDP80RO172OR000900070038-1
 Approved For Release 20051081Q% CIA-RDP80RW20R000900070038-1
and IC Staffs, odd jobs like mine)'than for. people doing
the real work of the Agency--collecting and analyzing
information. This, of course, is a more important
indicator to analysts of how important their trade is in
the Agency's scheme of things. The Estimates Board used
to be a way (admittedly much abused) of going to "the
top" (GS-18) and still being purely substantive. Now the
time comes quite early in successful careers when the only
way to go on advancing is to become a bureaucrats. Is this
consistent with your repeated statement that the ultimate
objective around here is substantive production?
3. OCI's own chain of command. Shorten it.. Some of
those layers have become places to put problem people- We
badly need a selection out process. Getting rid of the
lowest five percent is a start. But what about those very.
Able people who would never fall into the lowest-five, but
who at some point in their careers begin to coast (and
who might not if they felt a continuing need to prove
themselves)? And those who, because they've done one
job competently for a long time, get promoted above
their competence and become a problem?
4. The OPR succession. Put a first-class, proven
political analyst in that job--not a "manager," or someone
senior who's background is irrelevant but who happens to be
up for reassignment. This appointment will be an important
Approved For Release 2005/08/02 : CIA-RDP80R01720R000900070038-1
 Approved For
,,Release 2005/08/02 : CIA-RDP80RO17722OR000900070038-1
-4-
signal, especially to OPR but in OCI as well, of how
importantly you take analysis, and analysts.
Three things you've already started are. a big help.
1. Extended area trips, preferably with a research
project in mind. Aside from being, more fun than the-
.
former two-three week tourist-type look. at an area (and-
there's nothing wrong with fun), a real immersion in and
study of a problem is so clearly a sign of more serious
Agency interest in developing deeper analyst understanding
of his subject. Better still would be an expansion of the
presently very rare practice of putting analysts in
for tours--useful both for area education and
contact with policy types.
2. Some of the NIOs are working quite hard at putting
analysts in their area in touch with policy types at the
appropriate level (e.g., State country desks)--what
STAT
STAT
calls the "Indians to Indians" approach.
3. Your own meetings with analysts and general
demonstration of interest in substantive matters. (And
whatever happened to those weekly meetings with one NIO and. his
workers?) This is far and away the biggest improvement
during my 11 years in .the Agency. A sense that you read,
listen to analysts arguments, peddle the stuff around. town
and solicit ideas for what questions we can usefully
Approved For Release 2005/08/02 : CIA-RDP80RO172OR000900070038-1
 Approved For Release 2005/08/025: CIA-RDP80RQJJ20R000900070038-1
address, makesour product more relevant in-fact and is
a great morale booster out here. One cautionary note:
analysts are professionally trained to look behind what
"leaders" say to what they mean and do. So just pep.
talks from you (abut your appreciation of them and/or
devotion to analytical integrity), if not supported by
pretty concrete demonstrations that analysts are as
important in your scheme of things as managers and
customers men, will only add to the cynicism around. here.
Approved For Release 2005/08/02 : CIA-RDP80R01720R000900070038-1