NATIONAL INTELLIGENCE PRODUCTION

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP91M00696R000900020012-3
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 44
• Document Creation Date: December 19, 2016
• Document Release Date: September 10, 2004
• Sequence Number: 12
• Publication Date: March 23, 1976
• Content Type: MF

Approved For release 2005/07/28: CIA-RDP91 M0069~00900020012-3
MEMORANDUi~ FOR: Messrs. Proctor, Duckett, Knoche, Carver
Herewith a draft paper on production. It repre-
sents a met ding of the views to ~vhich the DCI has been
exposed, and of the.. suggestions far change that have been
put forward.
I am sendin a draft today for the DCI's
reading along wi a note explaining that
you have no ye seen it. I: .would like to have your pre-
liminary corxnents, written or oral; by noon Friday.
After the package has been refined, we will.schedule a
meeting or meetings with the DCI.
25X1
Ric ar a man
Date 23 March 1976
25X1
Approved For Release 2005/07/28.: C~fA-RDP91 M00696R000900020012-3
 Approved For Rise 2005/07/28 :CIA-RDP91M006961~0900020012=54_~
SUBJECT rational Intelligence Production
Background
7. In 1973 the Director of Central Intelligence abolished
the Qoard and Office of Pdational Estimates and set up the ~lational
Intelligence Officers in ti~eir place. This action began a debate
on the relative merits of the t',fo institutions that continues both
inside and outside the Comrnurrity. In 1976, as a result o~~ E.O.
11905, the DCI must again address the question of national esti-
mative intelligence. This m~rroranda~n analyzes the strengths and
weaknesses of each system as they ;rave emergAd from the debate,
and suggests ways which the stronger elements of each might ~~e
combined.
``~ 2. The Gffice of National Estimates consisted of a Board
of generalists supported by a drafting stafF of specialists. Thy
iJIO's are themselves area er functional specialists, grid are
charged with drawing their drafting assistance fro;n the analytic
elements of the Community. On ti;e other hand, the ~iIO's have
greater responsibilities than did 0/!`~E. Not only do they produce
a broader range of national intelligence Char. :lid 0/~~E, but they
also serve individually as ;t::f-F a`fi..ers 'co the DCI in t.+eir dress
of specialization.
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For F~se 2005/07/28: CIA-RDP91 M006961.~Q,Q2bU12-3
? 3. There are at least seven icl2ntifi:~bl? functions thai.
~~
have been perForrned by one or both of these r;sti~utions. E3oth
~?rere concerned t?ri th the management and revi-e~,r of national pro-
duction, a-td 0/NE witi~ production itself. Tt;~ iIIO's are directly
charged with consurner liaison, t?rith liaison Gc?ross Comm!~nity col-
lection and production, and ~,rith serving as an energizer of the
intelligence ef-Fort in new trays and directions; 0/i:E p]ayed a much
larger role in these activities in its earlier years than it did
in its later ones. The Board collectively advised the DCI on
major substantive issues; as noted, the PlIO's serve individually
as his substantive staff. The next few paragrap;rs discuss each
of these functions and how it has been handled.
Functions of a Pational Intelligence Staff
4. Consumer liaison: active measures to maintain close
contact ~?~ith the consurner; definition o-F his needs for intelIi-
yence; transmission of those needs to the production manager and
the analyst, including the policy context in ~?rnich they occur;
facilitating policy-intelligence dialogue; securing consumer reac-
tion to intelligence assessments.
Comment: There is .virtually uni-
versal agreement that the NIO's have been
successful in consurner liaison, far more
so than any previous e~Ffor?t. A ~:~uch closer
link has been establisnad betr;een the pro-
ducer and the consumer, and tyre product is
Approved For Release 2005/07/28'?EIA-RDP91 M00696R000900020012=3
 Approved For Fuse 2005/07/28: CIA-RDP91 M00696+0900020012-3
~,
not been any more systematic than their
predecessors, however, in general assess-
meat of consumer acceptance. O/~iE, par-
`a. - Cep ~ e ~ ~~ -"
Approved For Release 2005/07/28 ~ CIA-RDP91 M00696R000900020012-3; _ ~,`..~ dh, f{~}
Oval .~s. !~`MC
~1Q W~tu~f.. .
ticularly in its later years, t?~as severely
criticized for. an "ivory to~~rer" attitude.
It ~vaited for the tasks to come to it
and insisted on answering the questions
it thought worth answering; as a result,
few questions were asked and fetiver ans:vers
were relevant.
5. P-9anagement of production: development of terms of
reference; assignment of tasks; arrangements for asse t;bly, co-
ordination and review; editing and mechanical aspects of pro-
duction.
Comment: 0/NE managed production
,efficiently, largely because it ryas itsel;
responsible for the production it was man-
aging and because it had a structured system
for the process. The NIO's are less effic
This is partly because the range of pro-
duction tasks is broader, but it is also
because their arrangements tend to be ad hoc
and because these arrangements often run
athwart the lines of command of the pro-
~__._.
ducing agencies. The analyst must serve t:vo
 Approved For Fuse 2005/07/28: CIA-RDP91 M00696~0900020012-3
6. Production itsel~F: orga:;iza~tidn; ~ra-;ting; asse~ni~ly
and synthesis of contributions; defensz of dr?a-Fts under coordi
nation and review; preparation o'f resulting final drafts.
Comment: Under O/ire, production 4?!as
carried out by the 0/NC staff. ks noted
above, this was a relatively efficient process
but the staff was to some extent isolated
from and aloof from the working analysts.
The result 4vas a product that was s;:~ooth and
literate, but which sometimes did -~ot face
up to the facts or to the hard analytic
questions. Under the PlIO's, the production
is done by the analysts responsibTz. This
has paid great dividends in a product
soundly based on analysis and data, as
~vell as in analyst rnorale. ~1nal;~sts in all
disciplines feel the need to participate much
more directly in preparation of intelligence
assessments for senior policy officers. The
product tends, ho~r~ever, to be some~vf;at rouc,ner
than under the previous system.
7. Review:
a. Internal for: orgzniZation; coherence;
accuracy; completei;ess; vigor of argu-
meet; validity of judgrr.ent; responsive-
ness to stated requi re~-,~n t.
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28 :CIA-RDP91 M00696~0900020012-3
b. External for: relevance to policy
concerns; responsiveness to ques-
tions that should have been asi:ed;
broad topics not addressed; counter-
arguments not considered; ~?~isdorn
of judgment; the "experience fac-
tor"; general quality in the sense
of the factors listed under internal
~~~
C~-n:~...~~ (o~rvtr
.\ ``t w'T ~ ~ \~ t7-
.?
f tom` C.r t~
~~
duction manager at branch or division level
review.
Comment:
Internal revie~fr should be
integral to the manage-m-z-il--t~of production .
Under 0/NE, this was the case~~ Under the
NIO's, it is often not clear whether the NIO
or the production organization 'is responsible,
and the product suffers accordingly. The pro-
loses control of his analysts when they are
doing a task for an NIO. Such a paper will
receive considerably less ranagement attention
than would one for which the manager is fully
responsible. The NIO is usually toa busy and
sometimes not qua.___ 1 i-P~ie_d to give drafts the
intensive review they require. Tire PlIO's so-
lution is sometimes to draft a rwiesr~ r from
among the production managers. This results
in a further distortion of t~:e cha-in-of-cormand;
-~~~J :, -F4~
}Ge
QS l~e~E ~~
~~cn 1c
-5-
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28 :CIA-RDP91 M0069~0900020012-3
in the extreme case, it can E:eep a senior line
officer away from his assigned job for as much
as six ~ronths at a time.
The board of P;ational Estil~~ates ryas at
its strongest i n providing ex~~.:erna 1 nevi eav.
E3ecause it worked collegially, it ~?;as able
to provide the broadly-based cverview that
NIO's acting individually cannot. In additior~, ~ 1 ~~l w~~
`~4! ~1~{OVA/
it was able to contribute an additional dim~n- ~ h~~ r,.~Mr?~
wisdom. The NIO's, on the other ha~~d, have
been criticized for the unevenness of pro- ~ ~~~,~
a~ ~-~ ?,f
duc~ti on and for a fal l -off i n quality. I n Si'+~~~.~ ~ly o h
.,-----
effect, because of NIO specialization, the 1M
broader overvie;v can now be provided only by
the DCI himself.
8. The "stimulus" function: examination of the ration and
the world to identify new questions for intelligence; review of
intelligence product to identify tiveaknesses and gaps in collection
and production; identification of new sources, approaches, and
methodologies; advice to the DCI on how to deal evith the above.
Comment: 0/NE ryas more active in
energizing the intelligence apparatus in
the 1950's than in later years. This is
more a cri ti ci sm of the ;~~ay the system was
rnanned than of ti;e system itself. The
Approved For Release 2005/07/28 6 CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28: CIA-RDPg1 M00696~0900020012-3
performance of the NIO's in this function
has been spotty. Some have been active in
collection, and others in identifying gaps
in production. Generally, ho;vever, thz
absence of generalists among them and
their focus on short-range policy support
has meant~h~+ tha broader questions have
received less attention than narrow ones.
9. Community liaison: action across organizational,function-
al, and disciplinary lines to focus analysts and organizations on
priority problems; maintenance of cross-Community contact and
kno~r~ledge among intelligence offices in the same or related fields;
similar activities in relation to external expertise, private or
academic.
Comment: 0/NE, as a closed society,
was relatively inactive in Community liaison.
It drew on various organizations for con-
tributions (which it of -ten ignored) but
made little attempt to draw together
analysts of various agencies and disciplines
or collectors and producers. In sharp con-
trast, this has been an area where the PlIO's
have done outstandingly well. They have
made the individual analysts, and par-
ticularly those in obscure fields and
peripheral functions, feel much more a part
Approved For Release 200~~7/~~VeCIAn~~~~'~1Ve100~~b~~2~S~~~S~020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M0069~0900020012-3
' across functions and discipline,, and.
have brought the analysts thems^lves
more into the su5staritive debate on
estimative issues. In the past, this
debate tended to be conducted by rep-
resen~tati ves iti?~o or three echelons re-
moved from the ~vorking analyst.
10. Substantive staff for the UCI: advising him on consumer
needs and intelligence capabilities; ensuring tha'c his o:?tn sub-
stantive needs are met; representing him in certain ford; focusing
Community attention an special projects that cross functional lines;
trouble-shooting.
Comment: Except in a very limited
.sense, (e.g., on National Estimates) the
members of the Board of National Estimates
did not serve as advisers to the DCI. In
~~vV V \N'
(~~ ~~_`~ ,,~ -~ fact, until the NIO system was created,
F~
~~~t - ..
~~~ DCI's did ti?~ithout any substantive sta~Ff
G
C
~~ ~ ~-
- ~ bridging directorates, much less bridging
~' ~ ~,~C ~`' agerici es of the Community. The usefulness
~~ ~~ ~'
p- \~ ~ ~~,~ ~~~ of such a function was demonstrated by the
\~
~ ~ ~~'~ ~~~~y" ~~ ~ coordinating activities of the Speci a1
~~ ~~ao
~~
Assistant for Vietnamese Affairs, and it
was on that pattern that the staff role of
the individual tlIO's vas deigned. Here
again, they have made a substantial con-
Approved For Release 2005/07/28-CIA-RDP91 M00696R000900020012 3
 the process and the analysts at the other end more d;rPCt1v ;n
~~~_~ shaping. the product. It has also provided for the DCI valuable
~ ", w,c ~yiv ay~t,cn~ is ~na~ ~ t, ~nVOIV2S Zne consumers d't One end Of
R}~ ~'s~.rvices no-t previously available to him. On the other hand, while
11. It is clear from this review that the great strength
the product has been improved 'in the sense of relevance and in ad-
~~`~ dressing difficult analytic issues, it has suffered in other ways.
~0'~ The responsibility for review has been fragmented by the operation
Se~~,F~r
of the PlIO's across organizational ccmmand lines, and the advantages
~---
of a collegial revied~~ have been lost. Similarly,. while the PJIO's
have been effective as energizers of the Community on short-range
gVest~ions, their preoccupation tivith immediate policy support and
the lack of collegial interaction means less attention given to
the longer range.
12. It would therefore appear that the problem areas lie
in the relationships bet~veen the PlIO's and the line production
organizations, and in the absence of the broader view provided by
a collegial board. There are, however, additional problems that
are not directly brought out in the analysis above..
,~ ~, 1
v~~~'~pp?r~ve a 8R 0~89~R0~0900020
Approved For Lase 2005/07/28: CIA-RDP91 M0069~0900020012-3
tribution. It should be noted, hourever,
that their utility in this capacity d~-
pends to a great extent on the s~ray each
DCI chooses to clo business.
 Approved For Lase 2005/07/28 :CIA-RDP91 M0069~0900020012-3
since each acts as an individual, no common ~' ~ ve:j 1
Cba'~e~~vt'r 1~o~4MM.
discipline is imposed upon then. C
-- Production resources are not infinite.
llhen one tracks dorm through the structure, one
often finds that the-entire resources of the
NIO are so broad that no one in~.lividual can
do them all well. In fact, on the more active
accounts that involve lar?g2 elements of the
Community, it is difficult for an PdIO and an
assistant to do them at all. %The result has
been a great unevenness of performance. Each
tends to specialize in those -unctions to
t~~hich he is most comfortable. l~ioreover,
-- The duties assigned to the individual
CIA or DIA on a specialized subject of some ~~a-r.?iould consist of, perhaps, eight or nine
generalists chosen largely fror~ outside thz
Intelligence Community. It ti?~ould be advisory
to the pCI, and -?~ou1d review national pro-
. duction for him and provide the stimi:l us
function for the Community. The PiIO`s :~rould
continue in their present role; the Board
would be a level of review ~ir:posed above
them. The advan~agzs of such an arrangement
are that it provides the maximum of indepen-
dent advice to the DCI and. provides his product
with a certain prestige deriving from the rep-
utation of the Board and its T;embers. Such a
Board would also be useful to him in resisting
external pressures of a political or policy
nature. The practical disadvantages of this
approach, however, are very great. It inter-
poses another layer in an already complex pro-
cess. The Board's relations :?~ith the t'!IO`s
Approved For Release 2005/07/28 2-CIA-RDP91 M00696R000900020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M0069~0900020012-3
and with the production elements of the
Community ti,~oul d be di ~Ffi cul ~t to manage. In
effect, there t~rould be too many senior officers
answering to the DCI and stumbling over each
other to get at production resources. Fnather
practical consideration is that the NIO's al-
ready have more GS-18 positions than the entire
CIA production apparatus. Imposing a Board
in addition to tt~e NIO's would rake the system
top-heavy indeed.
-- A second approach is to create a
Board composed of the P1 I O' s . T hi s by far i s
-the simplest approach, in that it ti~~ould re-
quire only that the present group act,
collegially. on major papers. On the ofher
hand, it would not help much. The individual
NIO's are already too busy to pay a great
deal of attention to other NIO's papers.
Moreover, as specialists themselves, the
added insights they can provide are limited.
proposal fora greater
involvement of USIB in national estimates is
really a proposal that USIB be a Board of
National Estimates. In theory, this is an
attractive idea. In practice, USIB members
have neither the time nor, in some cases, the
Approved For Release 2005/07/281:3CIA-RDP91 M00696R000900020012-3
 Approved For fuse 2005/07/28: CIA-RDP91 M00696~0900020012-3
personal characteristics required fo?r
th~i s k i nd .of i ndependent revi etiv. The
sheer volume of national production is
curh ?i-hat rlotai7~rl ~..,..~,.?. ...~ .4~,,,. ~tc~rn
be made.
-- Yet another suggestion is for a
part-time Board consisting of the senior
officers at, say, the Office Chief level
in the various producing agencies. The
NIO system would continue as be-Fore, but
these officers as a group would be asked
to review a paper before it teas transmitted
to the DCI. This again is a relatively
simple fix, one that has the advantage
of involving the officers whose analysts
have prepared the paper and who themselves
are substantively capable of reviewing it.
On the other hand, this is hardly an in-
dependent reviea?1, and the production
Office Chiefs are busv mPn alr~arfv
~?
~~
V n - ~t
-14- ~ ~ ~
Approved For Release 2005/07/28: CIA-RDP91 M00696R~00~00020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M0069~00900020012-3
-- A fifth suggestion would be to charge
the NIO's with assembly of an appropriate
panel of revie~vers for each paper. This sug-
gestion has considerable i~~erit. The reviewers
could be chosen with regard to the content
of the paper, and could include senior officers
of the intelligence agencies, individual analysts
or outside experts. The disadvantage is that a
large-scale assembly of such panelists from
among officers ti~{ho have oth`?r commitments might
prove difficult, especially vrhen the novelty
wore off.
-- Finally, there is an idea of combining
a group of NIO's acting collegially with a few
generalists. Such a Board might combine the
best of the generalists' and specialists' worlds.
An independent review could be provided, and
.___ .there would be some Board members who could give
attention to longer-range and cross-disciplinary
problems. A disadvantage, ho~~rever, is that the
addition of generalists to the present dozen PJIO's
would make the Board unwieldy in size and, as
noted earlier, ,top-heavy in relation to the pro-
- duction structure. It would be nzcessary -to
reduce the number of NIO's serving on the Board
in order to accommodate generalists.
-15-
Approved For Release 2005/07/28: CIA-RDP91 M00696ROQ0900020012-3
 Approved For fuse 2005/07/28: CIA-RDP91 M00698~0900020012-3
15. Production hlanac~ment
-- One possib~il.ity, strongly advocated
by those who served in the Office of ~dational
Estimates and strongly opposed by virtually
everyone else, is the restoration of a draf-
ting staff similar to that 0/i~'E. The argu-
ments.in favor of this are largely that most
analysts do not have the required drafting
skills and all analysts are too busy analyzing
to give the proper attention to an estimate.
The disadvantages are that the system places
a layer between the analyst ~vho is at grips
with the problem and the national intelligence
product. It creates an element of elitism
and causes unnecessary resentment in the pro-
duction organizations. In addition, the exis-
tence of a drafting staff tends to create
drafts, whether they are needed or not.
-- Another solution, this one extreme,
is to make the production elements themselves
into a national drafting staff for the fu11-
range of national intelligence from the most
current to the longer-range estimative. In
practical terms, this 4vould involve redesig-
nating OCI as the Office of National Intel-
ligence and incorporating in it econo;~ic and
Approved For Release 2005/07~:-CIA-RDP91 M00696R000900020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M00698~0900020012-3
.military elements able to draw on th~~
research components of CIA, DIA, and State.
This Office of National Intelligznce t~rould,
to the extent possible, be jointly manned
by the three agencies. The NIO's ,vould
become the component chiefs within the
office. A chief advantage here 4vou1d be
the. consideration of one "national" office
under .the DCI of a large proportion of the
resources needed to meet national production
needs across their full range. A disadvantage
would be that it would sharply change the
character of the NIO's; it ~rould improve their
ability to manage production, where there are
now problems, but at the cos t. of wea!~ening them
in what they do best, if only from the demands
on their time. The change would 6e disruptive,
C1A
particularly for the structure of and it
would be difficult to maintain smooth relations
between the Office of National Intelligence and
the Community research organizations Upon which
it would have to depend.
-- In contrast, the least radical change
would be simply to add additional assistants
fora fear of the busier ~1I0's, or a small pool
-17-
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For fuse 2005/07/28: CIA-RDP91 M00696~0900020012-3
of generalist drafters. 6~hi 1 e this Lvould
re lieve some of the pressures on the syster~,
and in soave cases provide better drafts, it
Nrould not help in any ~?ray ?to-resolve the con-'
fl~icts 6ettiveen the NIO system and the pra-
ductior, manager. Rather, it would be seen `
;.
as the first step back toLVard an 0/fiE-style
drafting staff and would be bitterly r2sis-ted
by the production o-Ffices. In fact, once 'the
line tivas breached, it ?rould be very difficult
to resist pressures for future additions_
-- It would also be possible to return
to the systern that tivas used for the mi 1 i tary
estimates in the last years of 0/i1E. Under
this system, the NIO's would request the DDI
and DDS&T to prepare a draft, but tlould not
6e responsible for its production. Rather,
the production offices would be responsible
for organizing Community inputs for drafting
and for Community coordination. This system
worked well, and solved the problem of keeping
responsibility for the quality of production
clearly in the hands of the production manager,
It would be seen today, ho,?i=ever, as a step
back toward CIA domination of tl:e production
process.
NO
-18-
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 -- Lastly, the previous suggestion could
be adapted to broaden the Community's role.
Again, the NIO's 4~;ould give up the production .
management function. A "Piational Intelligence
Steering Group" t~;ould be established under NFIB,
chaired by the DDI and. ;;rith hi s opposite trembers
from DIA and INR as members. This Committee
would be responsible for production management.
It would allocate tasJcs and develop terms of
reference in consultation with the PlIO's, but
it would be responsible to the DCI ror the quality
of .the product. Such a device ~;rould lessen the
burden on the NIO's, leaving them free to do tihat
they do best, while placing the responsibility
far production clearly s~r-ithin the chain-of-command.
The disadvantages are that the Steering Group
becomes another layer in the process, and that
it/might place inordinate demands on the time
its membership.
C~
CvN~vn~ .
16. Changes in -the NIO's themselves. Implicit in a number
of the suggestions above are char~g~s in the character of the i~IO
system. Some of these could be made independently of other changes,
ho~;ever, and it seems useful to analyze th~e,n separately.
-- A suggestion has be?n made simply
to abolish the system. Sentiment through-
Approved For fuse 2005/07/28: CIA-RDP91 M00698~0900020012-3
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For ~se 2005/07/28: CIA-RDP91 M00696~0900020012-3
out the Community, however, is over-
tvheln~ingly in favor of its retention in
some form, and there seems little need to
discuss this further.
-- A more realistic idea, and one
that is built into some of the suggestions
for production management, is that of re-
lieving the NIO's of this management func-
tion. As noted earlier, this would give
the NIO's more time for their staff re-
sponsibilities to the DCI and their con-~
Sumer and Community liaison functions.
It would relieve the pressure to provide
them more staff and rnight make it possible
to reduce their number. hlost important, it
woul d resolve the conflict bets~reen them
and the line production managers. On the
other hand, it would vitiate the original
concept of the PlIO's, and r~iyh-t weaken their
ability to serve the DCI.
-- It has been pointed out that some
NIO's are much busier -than others, that Borne
have broad Community resbonsibilities rrrhile
others are more narrowly focused and have
few Community resources to support them.
Thi s raises the possi bi 1 i ty that sor.:e !~lIO's
Approved For Release 2005/07/~~; CIA-RDP91 M00696R0009000
 Approved For Lase 2005/07/28: CIA-RDP91 M00698~0900020012-3
should be full-time and others not. Coupled
with this is the thought that some generalists
might then be added to the ~JIO group.: Part-
time NIO's would combine their PJIO responsi-
bilities with management of appropriate pro-
duction elements. In practice,~ho~vever, these
would have to be CIA officers, and the Community
aspect of the hI0's would be correspondingly
weakened.
-- Finally, it would be possible to go
one step further and make all the MO's-heads
of~corresponding production organizations. If
these organizations .are those of CIA, this
suggestion amounts to abolition of the system
and return to the concept of CIA as the DCI's
coordinating Staff. If they are those of the
Office of Pdationa7 Intelligence proposed above,
-- the Community aspect is retained, but NIO's who
are organization managers will have little time
for their staff or liaison roles.
Some Possible Approaches
17. On the basis of this analysis, it is now possible to
assernble the parts into models of tvoring systems. 41e have developed
five such models, rangi~;g from something resembling the present PdIO
-21-
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M0069~00900020012-3
system to a rather radical change in the tiday the Commur-ity does
business. These rnodels are illustrative only; features of one could,
in many cases, be adapted to another. For si~r:plicity, the diagrams
emphasize the production of National Estimates: A separate set
shows hood current intelligence would be produced.
18. P~;odel I is the present system wits minor modifications.
An external review process t~rould be added either by requiring a
collegial revietid of estimates by several P~IO's or by the formation
of ad hoc review panels. Three positions for generalists recruited
externally might be created by elimina~tion~of the NIO's for Special
Projects (.however useful, is this an NIO function?) and for Economics
(duplicates D/OER) and a consolidation of five assigrjrnents to pro-
vide one NIO for East Asia (including Southeast), one for t~tear East
and South Asia, and one for A-Frica and Latin F.merica: Each of ~che
three might be provided an additional assistant and the generalists
an assistant apiece 4dithout increasing the size of the itiIO organi-
''nation.
19. Model I would require little disruption of present
activities. It would meet some of the felt treaknesses o- the NIO
system, especially in regard to external revi~;d and broad projection.
It would establish.a better balance among htIO responsibilities. It
would not, however, do anything about th ~ce~ problem o f d i vi ded ~j ~'' f1~'1~"r/c
authority. ~-~. ~' ~ r ~ Y ~~ w,~y~
20. Model I I addresses i ?tsel -i speci ~i cal ly to thi s problem. ~~ Ft f'~,~t~~
The NIO's would be relieved of any responsibility for production
Approved For Release 2005/07T2~-CIA-RDP91 M00696R000900020012-3
 Approved For Lase 2005/07/28: CIA-RDP91 M00696~0900020012-3
management. This responsibility would be placed in a sub-committee
of NFIB consisting of CIA/DDI, Chairman, and corresponding DIA and
INR members. All requests for inter-agency production would be
addressed to this Steering Group, and it ~rould decide trhether the
task was to be undertaken, by whom it aras to be drafted, and ho~v it
tvas to be managed and coordinated. Terms of reference mould be
worked out in consultation with the appropriate P~IO. Thus line
managers Hrould again become fully responsible for their product.
At the same time the valuable staff and liaison services performed
by the NIO's would continue.. The present "production committees"
2l. Under this arrangement the NIO's might be reorganized
as in Model I, but with a reduced supporting staff. Their total.
number might be reduced.by further consolidation or by a system
under which only those with major Community coordinating roles
of USIB would report to the Steering Group.
(Strategic Programs, Conventional Forces, USSR, China; C~liddle East)
wcu kt:.be full-time. CIA officers in appropriate substantive line
positions would assume the duties of the others.
22. P~odel II meets most criticisms of the present system
and provides the means fora significantly better product. It
preserves the best factors of the NIO system, notably a substantive
staff directly under the DCI to monitor the full range of Community
~4~ 1_rsc{u~~~~dM p_l~~ wk.~ _ww
,~ .
e
n
Approved For Release 2005/07/28: CIA-R[~Ft~61~~1p(~9~R0(~g9~l~q,~~j2~3~~
activities. Its major disadvantage would appear to be uncertainty
as to whether the NIO-Steering Group-producing element linkages
would work
f~icie
tl~
 Approved For ~ase 2005/07/28: CIA-RDP91 M0069?00900020012-3
? 23. Model III is the simplest and most efficient of those
presented. It would attach to the DCI an advisory Board (o-- PIIO's,
of generalists, of a combination), but return to CIA the production
management and coordinating function.
24. As compared with t?1ode1 II, Model III would provide
roughly the same degree of review. The "Board" could be consti-
tuted in any of the ~vays suggested.. The NIO`s could be retained
and could do everything they noUr do except production management.
The change would-place the basic responsibility for drafting ~~rith
~~
those elements dedicated exclusively to national intelligence. ~~h-~t~;~~~~~
This would in general produce better drafts, but it would also ~+-t~ct1+}tucc~~~~
mean that CIA would be preparing them. The DCI t;ould have to be ~M~~a"'~
~b~ttl~~~ate~ 9~,
t~ri 11 i ng to defend the concept that CIA i s his staff if he tirants to
take thi s route.
25. .Model IV is the BoardJStaff of National Estimates_as it
used to operate. The Senate Select Committee apparently grants a
return to this structure. It ~vould be very difficult to do so,
especially after analysts throughout the Community have been more
deeply involved in production. The consensus is that there are
advantages.to a collegial board but that the staff has been over-
taken by the maturing of line analysts. This solution, as noted
above, no longer is a valid one.
26. Model V is the radical solution. In it CIA's Office
of Current Intelligence would be used as the frameti~rork for an Office
of National Intelligence anstrzring to the DCI. Such an office,
jointly managed and manned by CIA, DIA, and Ii?~R, Vrould be responsible
-24-
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28: CIA-RDP91 M00696~0900020012-3
for? the ful l range of i ntel 1 igence pr~o:luction for pol ~i cy support,
from the most current intelligence. to national estimates. It
would be organized regionally, !?rith the bull: of its analysts
political, but each regional division !?rould have strong economic
and military staffs. These ~vould in turn drag on the research
elements of DIA, CIA and INR. The latter ~?rould no longer deal
with current support matters, and the new office would be solely
responsible, not only to the DCI, but also to the Secretaries of
State and Defense. ?
27. A generalist t3oard of Review might we71 continue ad-
visory to the DCI, but under th~?concept the PdIO's s?rould become
the line officers responsible for production in their areas of
interest, e.g., the Chief, i~estern Hemisphere, of the Office of
National Intelligence would also be the NIO for the 4)estern Hemis-
phere.
28. Model V would be a real departure in the direction of
Congress' original inten~ in 1947. It, in effect, reinvents CIA.
~~.
It wauld__also be extremely difficult to administer vreli. In par,
ticular, joint manning could be a problem. The core t~aould have to
be CIA, but neither DIA nor State would readily come forlrard !vith
personnel of comparable quality. If they did not, and if the
organization could not be managed as a cooperative enterprise, then
the DCI would have given up effective CIA mechanisms for even more
cumbersome Community coordination.
29. On the other hand, P~1ode1 V ~?rculd provide a single source
for intelligence in support of national policy. Ti~us the same
analysts, using the same sources, arould be producing current intel-
Approved For Release 2005/07/~~CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28: CIA-RDP91 M00698~0900020012-3
ligence, national estimates, and all the forms in betareen. This
vrould provide a greater coherence to national reportiny. P?todel V
-?rould provide the maximum centralization of the production function
under the DCI . I t also ~?roul d have the di sadvantayes o~f ~%~odel IV
in spades: potential elitism, isolation of aralysts frcm drafters,
and drafters from hard facts; NIO's limited in their cross-functional'
roles.
Approved For Release 2005/07/2~5CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
Estimative Intelligence Estimative Intelligence
Model II
Motlel I Present system with production management
Present system with minor modifications separated from NIO's
DCI
Board of NIO's
NFIB
National
Intelligence
Steering Group
CIA, DIA, INR
DIA
CIA
N 10's
INR
DIA
DCI
NFIB
CIA
DRAFTS
~' ADVICE & CONSULTATION
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
INR
 Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
Estimative Intelligence
Model III
Adaptation of system used for military estimates 1970-73
DCI
NFIB
CIA
DCI
NFIB
Board of
National
Estimates
?
DIA
Board of
N 10's
INR
' DRAFTS
e..r ADVICE & CONSULTATION
CONTRIBUTIONS
Estimative Intelligence
Model IV
System in use 1951-69
DIA
CIA
INR
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Releas Q5/07 28
: IA- P91 M00696R000900020012-3
s Ima Ive
Model V
me Igence
Central community~acilit\~roductio~
DCI
NFIB
Office of
National
Intelligence
(Jointly manned)
__
.I
l
1
esearch onl I
~ i--------~= ----------+
~ I' ~ I
I ~ 1~
DIA
CIA
INR
DRAFTS
ADVICE & CONSULTATION
~' CONTRIBUTIONS
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
Current Intelligence
Model I - IV Model V
DCI
DCI
DIA
CIA
`~ DRAFTS
INR
CONTRIBUTIONS
DIA
CIA
Otfice of
,National
Intelligence
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Release 200547/28 : CIA-RDP91 M006:96R000900?20012-3
OPTION 1
DCI
.CFI
DC I`: CHR.
COMMUNITY MGT
NATIONAL INTELLIGENCE
CIA MGT
Other ~Rrod
Approved For F~efease-2005/07/28, : CIA-RDP91 M00696R000900020012-3
PRODUCTTON-EVALUATION NT FAC
NFIB
DCLCHR
??DC 1~i4 -~ ~ H
~f/~ i~1
CXC' Scr a
A BAN c}~
-~.~.
~h~
 ILLEGIB gpproved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 THE DECISI,~,$Ed1~~se 2005/07/28 :CIA-RDP91 M00696~0900020012-3
CFI HANDLES RESOURCES
AND MANAGEMENT
OPTION I
ADAPTATION
OF PRESENT SYSTEM
I
I
I -
L----
HOW TO ALLOCATE MISC. FUNCTIONS
AND ORGANIZE DCI STAFF (IG, OLC, ETC.)
1. THE PRESIDENT HAS DECIDED BY
EXECUTIVE ORDER
DCI RESPONSIBLE FOR
EVERYTHING ELSE
2. DCI NOW CAN ALLOCATE NATIONAL
PRODUCTION FUNCTIONS AMONG
HIS SUBORDINATES
OPTION II
NATIONAL BOARD
OPTION III
NATIONAL PRODUCTION
UNDER
AGENCY DEPUTY
HOW TO ORGANIZE NATIONAL PRODUCTION
MANAGEMENT
596668 Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Fuse 2005/07/28: CIA-RDP91 M00696+0900020012-3
Location o-f the rJational Production Function
- The decision to be taken no~~v is only on location, not
how the system is to 4,rork (NIOs, Board of Estimates, etc.).
No ~?reight should therefore be given to the words (NIO, etc.)
in the production box.
- However, a decision on location does have organizational
consequences at lower levels, and these must be considered in
making it.
- To this end, the Option charts show some of the logical
consequences. These are illustrative, however, and certain
features of one could be adapted to another.
- Organizational elements not relevant to this problem
are not shown. No attempt is made to show the organization of
the IC Staff. The boxes represent generalized functions only.
One major question is the l7SIB Committees; -it is quite possible
to visualize the production Committees (EIC, etc.) hanging on
NFIB, and the collection Committees (CO~IIREX, etc) hanging on
CFI, but it is premature to face this question. The "NFIB Sec-
retariat" function therefore remains to be defined.
- None of these Options preclude seine port of personal
substantive staff for the DCI.
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
Option I
- Sets up Community Deputy as .responsible for all Community
matters except production .
- Creates National Foreign Intelligence Board, advisory to
DCI, with Agency Deputy as Vice-Cha rman. Community
Deputy is responsible for support of PdFIB and its Com-
mittees
- Fakes Community Deputy a member of NFIB
- Continues NIOs in their present role
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 OPTION I
Deputies Divided
Community-Agency
National Production Separate
CFI
DCI Chm
DDCIIC V-Ch
DDCIIC Exec Secy
IC STAFF
CFI Support
NFIB Secretariat
COMMUNITY MANAGEMENT
NATIONAL INTEWGENCE
CIA MANAGEMENT
--- PRODUCTION-EVALUATION INTERFACE
NFIB
DCI Chm
DDCIIA V-Chm
DDCIIC Memher
DDCIIC Exec Secy
596665
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07C28~ ~`. CIA=RDP91 M006'96Rb00900020012-3
Option I
PRO's
- National intelligence given strong emphasis
- DGI's Community role clearly defined
- DCI has close access to substance
CON's
- NIOs must rely,on, and will be viewed
as, an element of CIA
- The affairs of al] three subordinates
of the DCI are entangled
- Disruption of CIA chain-of-command
- Some do4vngrading of CIA
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
Option II
- Replaces NIOs with a Board of National Intelligence, chaired
by the Agency Deputy but reporting independently to the. DCI
- Makes Agency Deputy Vice-Chairman and Executive Secretary of
NFIB. Community Deputy remains a member
- Transfer NFIB support responsibility to DDCI/A
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
 Opproved For Rel~se 2005ff17/28:CIA-RDP91 M00696R000900020012-3
OPTION II
National Board
CFI
DCI, Chm
DDCIIC V-Chm
DDCIIC Exec Secy
DCI
Board of National Intelligence
DDCIIA Chm
DDCIICM
CFI Support
COMMUNITY MANAGEMENT OTHER PRODUCERS
NATIONAL INTELLIGENCE
CIA MANAGEMENT
- -PRODUCTION- EVALUATION INTERFACE
NFIB
Secretariat
NFIB
DCI, Chm
DDCIIA V-Ch
DDCII C Member
DDCIIA Exec Secy
DDI
DDS & T
596666
Approved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
Option II
PRO's
- Maximum emphasis on national production
- Community Deputy no longer tangled in Agency
Deputy's business
- GCI`s Community role distinguished from
Agency
- DCI has close access to substance
- CIA's primary role better recognized
CON's
- Dual role of Agency Deputy administra-
ti vely awks~rard
- His association with Board somewhat
obscures its ecumenical image
- CIA chain-of-command still affected
Approved For Release 2005/07/28 :.CIA-RDP91 M00696R000900020012-3
 Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
Option III
- Agency Deputy becomes Deputy for National Intelligence and
is responsible for day-to-day management of Agency
- Divides DCI's responsibilities between his two Deputies on
the general basis of resources to one, substance to the
other
- Recognizes institutionally DCI's need to be impartial in
resource matters and his close ]ink to CIA in substantive
matters
- NIO successor organization placed under DDCI/NI .
- NFIa arrangements same as in II
Approved For Release 2005/07/28 :.CIA-RDP91 M00696R000900020012-3
 OPTION III
Deputies divide~'pproved For Release 2005/07/28 :CIA-RDP91 M00696R000900020012-3
Management-Substance DCI
CFI
DCI Chm
DDCIICM, V-Chm
DDCIICM, Exec Secy
IC Staff
CFI Support
NIO
or Successor
COMMUNITY MANAGEMENT
NATIONAL INTEWGENCE
CIA MANAGEMENT
PRODUCTION-EVALUATION
INTERFACE
596667
NFIB
DCI, Chm
DDCI/ NI V-Ch
DDCIICM Member
DDCIINI, Exec Secy
NFIB
Secretariat
AGENCY
DDS & T
pprove or a ease - R6009tj0020t~
 Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3
Option III
PRO's
- National intelligence continues primary
- CIA role recognized
- Balanced responsibilities of deputies and
clear separation of their roles
- Makes possible clear chain-of-command for
production
CON's
- Not as Community-oriented as other Options
- Can be criticized as "CIA in disguise"
- DCI one step further removed from production
- DDCI/NI will give less attention to. Agency
management
~~
Approved For Release 2005/07/28: CIA-RDP91 M00696R000900020012-3