MINUTES OF THE 14 APRIL 1977 EXECUTIVE ADVISORY GROUP MEETING

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP91M00696R000200010010-3
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 25
• Document Creation Date: December 15, 2016
• Document Release Date: February 26, 2004
• Sequence Number: 10
• Publication Date: April 26, 1977
• Content Type: MF

~.v ~ n v a 4? ~~na
Approved For`Release 2004/05/13: CIA-RDP91 M006 6 22QDDA 0010-3
I
Copy 4
MEMORANDUM FOR: Executive Advisory Group Members
SUBJECT Minutes of the 14 April 1977 Executive
Advisory Group Meeting
1. The Executive Advisory Group (EAG) met on
14 April 1977 to discuss a proposed memorandum to the
DCI, entitled "CIA Views on the Future Management of
the Intelligence Community" (EAG 22), drafted by the
Comptroller as a CIA contribution to current deliberations
on Presidential Review Memorandum NSC-ll, which directed.
a review of the organizational structure of the Intel-
ligence Community.
2. In sum, the paper recommended that the DCI seek
statutory line management authority over the National-
Security Agency and the National Reconnaissance Office
in order to allow him to exercise his responsibility
for management of the Intelligence Community.
3. Discussion centered on the paper's neglect
of the DCI's role as intelligence advisor to the President,
which competes for his time in his capacity as manager
of the Intelligence Community. It was agreed that the
paper should be redrafted to take this role into account
and to define other options which would involve less
radical change.
4. Subsequent to the meeting, the paper was
redrafted, as requested, and circulated as EAG 22/a to
EAG members for comment. It was delivered to the DDCI on
22 April for transmittal to the DCI.
25X1
25X1
Secretary
Executive Advisory Group
25X1
cc: IG
LC
Dep/NIO
Approve
b'SFor Release 2004/05/11W-' Oq,,FDP91 M00696R00
 Approved For`Telease 2004/05/13: CIA-RDP91 M0069000200010010-3
D D
R R
A A
F F
T T
MEMORANDUM FOR: Comptroller
SUBJECT : Another Option for the DCI
1. The discussion last night made it clear that the DCI cannot
Q c.,.. __~, J " aa,.Ta.;..
go back to the President wit A e status quo. The President wants to
see a substantial increase in the DCI's authority. On the other hand,
a number of objections were presented to the solution that you proposed,
all of which have to be given some weight.
-- The DCI would have to be so much a manager
that he wouldhasie to neglect his principal job as
intelligence adviser to the President.
-- CIA would be fragmented and ultimately
swallowed up in,NSAA and NRO. (Some of this
sentiment- - 'lts-f --' the emotional resistance of
those of us who are long-time CIA officers to such
a concept, but some of it reflects the obvious
losses from breaking up a piece of effective and
integrated machinery.)
-- Such a move would not be politically
feasible in view of DoD resistance.
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
Secret
 Approved FoMelease 2004/05/13 : CIA-RDP91 M0062000200010010-3
2. It seems to ,,,e-that much of the argument against your proposal
liter
derives from the model that everyone was assuming last night. G,c
c& that the relationship betwpen the DCI and the D/CIA, D/NSA,
t
and D/NRO would be like the present relationship between the DCI and
A
DDCI. If we use another model, one in which these three relationships
were similar to that now existing between, say, the D/NSA and SecDef,
these arguments would have less force.
3. There are two basic for the creation of a stronger
DCI, better intelligence and more efficient management. We and the Senate
Select Committee place more weight on the former; OMB and House Appro-
priations focus on the latter; the President wants and the country
deserves both. For us, at least, the key question is: How do we get
better intelligence?
4. To begin with, we must emphasize the DCI's role as the Presi-.
dent's substantive intelligence adviser. That in turn requires that the
&A11 614
DCI have independent intelligence production capabilitam under his con-
trol and the time to wen-with nd shape the output of ai9_R4.S!& to meet
presidential and other national requirements. Such a DCI _canno-t~spend the
bulk of his time either on management and resource problems or on fighting
fires stirred up by the Congress, the press, and the Department of Justice.
Such a DCI, however, needs to have the major collection systems immediately
responsive to the requirements of his production organization. Over time
it has become clear that many~of these systems, particularly those in NSA,
are somewhat ess than responsive to his requirements and that all of them
can only be brought to respond through cl abor ter., bewildering, and time-
consuming collegial procedures. Moreover, the lack of central authority
Approved For Release 2004/05/13 _,QIA-RDP91 M00696R000200010010-3
 Approved Foelease 2004/05/13 : CIA-RDP91 M0062000200010010-3
has meant that the case for the development of certain collection
capabilities clearly needed to tmeet important analytic has not
been effectively made either to Congress or to the OMB, vii
Finally, the anomalous position of the DDCI ..as the primary
advocateAprograms d to support ttr;-. national intelligence production
w~+r
-~api].ity has, especially E.O. 11905, prevented him from arguing his
A
case as vigorously as its intrinsic merits deserve.
5. If one follows this chain of reasoning one ultimately reaches
the same t as those advocating stronger nagement of the Community
for reasons of budgetary efficiency, i.e., the DCI should have as much
authority over the other two major national programs as he does over CIA,
lass -.i' _ sar . We submit,
as did your original proposal, that this be linethority not just
budget, for the reasons t At you cite and als?1`- prow de"arrangements
under which the DCI can fulfill his responsibilities for maintaining
A
legality and propriety. However, because we would put his substantive
responsibilities first, we would not move in any major way toward
centralized administration of the three programs. Rather, at the first
stage we would establish a DCI with aAsmall staff and place under him.
three statutorily established separate agencies. Their directors would
report to him and their budgets would be allocated to him. But their
directors would be responsible for the management and administration of
their agencies. The Directorate of Intelligence would remain within the
Central Intelligence Agency for purposes of management and administration,
but the Deputy Director for Intelligence would report directly to the DCI
I'
on substantive matters. (ae will not attempt to address here the relation-
25X1
Approved For Release 2004/05/13-.3CIA-RDP91 M00696R000200010010-3
 SECRET
Approved FoCelease 2004/05/13 : CIA-RDP91 M0062000200010010-3
ship of the present NIOs to the DDI, although we recognize that this
proposal ' a closer integration than at?4'present. Neither do 0
address the question whether the DDI should develop across-the-board
analytic capabilities;t this must ultimately be a4u~.11 as well as
the even t lrnier question where the resources for a stronger DDI might
be found.) 6. Such a solution would.oombine a DCI not overly burdened with
management t irL : l i? h ? e--# . and, tfre capabil
produc ion an1wia the authorities necessary to ensure that collection
- Q
serve those capabilies properly. It preserve)( the integrity of CIA
and the obvious benefits tht And, because in this
j,,,~,cT ? C,s~?
first stage NRO and NSA remain separate, it"ss reversible,either if the
arrangement prove a failure or in the event of war. This last would
make it at least marginally more palatable to the DoD. Moreover, it is
a real change, and one -Weh should satisfy the President's desire for
centralized authority. As opposed to your original proposal, it would
not go so far toward efficient centralized management, but the preser-
vation of the unique qualms and strengths of CIA seem to-us worth
this sauce. Overall, it place, relatively more weight on the DCI as
4 r.1 4.e"
substantive adviser to the President and relatively less on ..his
[}Par'
rues: alrl~vtiredsAt a later
stage/after the dust ha& settled and after the DoD 4 persuaded that the
_tk.
detachment of NSA and NRO ha been accomplished without reducing intel-
ligence suppor rationalization of the various collection
-4-
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 SECRET
Approved Foi"'lCelease 2004/05/13 : CIA-RDP91 M0069000200010010-3
capabilities under the DCI might be undertaken, perhaps ultimately
leading to the creation of a deputy to the DCI for national intel-
ligence collection. But in +.u view, that can wait.
Approved For Release 2004/05/13 : t A-RDP91 M00696R000200010010-3
 ANA
Approved Fo elease 2004/05/13. CIA-RDP91 M006
000200010 -3 r
:
.CI'l C~.C! l' C' s.,; ~L ).iC-y 'f:(1^ Ul: ^ S :l. C) 7T UA. i:'"?:. 0:,~: e ~! ?C:4?~~rn
o_ c?Ilizat1C71?cal arrang'~?] e l't~S :Cor T l^ '_r 't:'2C i?~ _.'_.:i ?' 5 1fl ?1 1C :flCC
T _. ? r ,,. L L n that goo: 1 :l: eZ ? is ~?nce is
:: n,2s it ~i v 1T~ Jo.- ~.C 1 4 4CJ underscore .r
cI ~] ime 1CqU-~ YE.ii ..:`ilt at every .I-eve- C; C+7 1 '~i:; =.^ r . concerned with
tC"`l u~ se cur ?_ty from th's Py.-C-osiddc ilt a:? d':_ iar S o ~.!? ii ? t? anal
}
Security Council to tbe
At the national level the pu'ri^'J:ies of i.:? a U. S . Il'L e ? 1C ~. CE' . .
C7:i itnity is to produce high Clual L ~.z', rel ~!c :t: r and objective
. dent - ~ tI?C 'SC ~ln li? e s n ;1.t for
.n clli gonce ox the Pre~s_Lc.
C i3 Cong ess. These national. needs J Lr t iO I ana
c *lcll sis sIzp poring the o_''rau? a tio=? o
providing s t3;ateg--c and tactics'. t ll~C?il;:c~
-co 1
s
the :..ib tec r. )Io r. ., , }' C ~":.;Atcad '?"C'C__d.~Lca
of -
the
xT!'t Cl.I-:i CiZ!-_1Cv also SC13.:' t. 'G.
r
~?C se a, rL ~1 _I .} Ate. onr level
.~~~,~~?:ct.r?,~~I.f:. of i1 t.1_~^'ri.~ an ~._L 3
?.i!tL ? l '- 1_S use in U ? S ::_'.'=7= X..-ea-0-As `J:'1~'i S SAC:;=??
t. iger_c le Pt a f: i r] ~ C% r c 3 -C5 a
to develop as well as in fo:ccc st_r_uct_.u-re plL::='_-^"q
'e r-c^r_?? t-i =11 Li'_J;C)5 1 : ~_Ofl S_C]ti,
Zc:VOjy C:J''..r s3-s
X t once c'.IIC~ support fo the con i?C i_ o? operations
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 1 iY :l udl t'i, ti 11.?_ urj L da'La on I it 11'. r;j orcc
a r, is i v i ,Appr fired tfr-x 1094e t2.OO4fO'S/~ : aAIRDI -911'DVbOd6}4 6bo 1~brb9 0-3
producing such intelligence are as diverse as are the t1O2d;.
I,, isstu? is !h 1`. oreanizationat-1 will
e+ i ecti yel y See ve the wi da Variety of intelligence rte---'is Of
national, departrtenta1, and tactical users.
The national Security Act of 19217 and the National Security -
Council Directives of the late 1940's and 1950's established the
basic division of responsibilities with respect. to intelligence
acts V1 tl es among agencies and departmen-ts . These divisions had
their origins in the traditional distinction betmeen military and
non-military intelligence. The CIA was directed to produce
national" intelligence while the mil -i -ta ry services were asked
to Continue to collect and pro 'uce' "military" 1nte11igerlce for
use both in war and peacetiii. `.
I n the charge to the Director of Central Intel i i ericr and :r
the 1947 act. -For. "of the intelligence activities of
the several Goveril:':ient depaartf ants," Presi derit 7 rti a an sought to
prevent a repeat of the intelligence con-Fusi on and delays that
occurred prior to Pearl harbor. The problem addressc:d under the
act Was ho,?i to collect, collate, and -process intelligence reports
and estimates that t iould best serve the national leadership--the
President and the NSC. -
Approved For Release 2004/05/13 CIA-RDP91 M00696R000200010010-3
 Iti; iS dear that 11. 11 e' :1 aGiona! s'ec.'r~i ty of the'
19 C's A rg ~ IFP' CRf ~a O~ tOt~ 13LilC[+'i p91jYOP6 002AR410010-3
to hno1 o_li cr i ly cu f tp1 ex i n"cell i gence cc,'. r.mi Ly. C e o1 d
distinctions I)etwe en "fr.! 1 i tary" and n?1n-~ it i t [ arj',' Lac ti cal"
and national have blurred.
There are two sepZrc; i e aspects to th ' probl e s
arisen since 1947: the efficient d rag^men t of the resources
allocated for intelligence collection and analysis; and the
Optimal targeting of intelligence assets assigned for purposes
of collection.
Resource Allocation has been co:i~pl I cated by the
i ntel I i g Lice e:ezi cli
huge increase in the proportion of the overall
is assigned to Signals Intelligence and to Satellite Reconnaissance
systems. I n turn, the portion o F the In tel l i ceilce Co; ~ mi ty `s budget
which is appropriated to the Director o: Central lnte11-cjezlce (in
his capacity as Director of CIA) has S.'i':~unk to This
~ci s ;rig o :'oral l
Co:f'.pl i cate d the problems of the C.' in ex-
,? r e t.is .of
r'.aii~gef end or, f the Intel l i gr`nce Co.I? Giiti 4, , and of USA and URO in, in effect, Ski the Sc ;i'.' 't.ime,
the Proliferation of intelligence collection cap abilities and the
increase in their costs make it increasingly irll;po;'an'- to consider
tradeoffs be twe en differing systei ,s .
Operational "l-asking, has been conpl icatec4 because
increasingly, intelligence data--collection systems have ;ro:r.i
capable of serving both the broad interests of the r:ol ; cy-rra'.:ers
Approved For Release. 2004/05/13 : CIA-RDP91 MO0.696R000200010010-3_
 a".)d de i"ense pl allll( and th? more spec) i i c techlll ( inter_e:i Ls
Approved For`Release 2004/05/13 : CIA-RDP91 M0069000200010010-3
of weapons developers and field co;''.mmanders. Signals intellige'r,ce
i; -0vi Gies ho th econolni c and Cili i i Lary Cori:;:luili ca t i ons i ri.:el g;ence .
Spies are asked to collect information on Soviet weapon technology,
political intentions, grain harvests, etC. Satellites produce
pic Gores which are critical bo'th to Lhe SALT policy--raker
Arry Commander on the East Ger~mlan border.
There are issues of ho,; to provide the tactical
and the
Commander
ill the field not only the appropriate product of "na'ii onal "
Intel 1 i gence collection assets, but ho,,; to pemlit that commander
to task; those assets directly to his needs; and there are issues
i n the opposite direction, though mainly Of enstiri ng that the
appropriate product of "tactical" intelligence collection is mad
available to national pol i cy-makers .
One key Cluest"IOM, then, is what degree 0" ccen ral i zcati on of
control in the Intel 1 i gence Cole; ?unii ty will provide resource ration,
al i nation and insurance against duplication and 1'ras te. Another
question is ho':, to task til rlultii:l-icity of collection systerns in
ways that w i l l be fully responsive to the needs of all consumers.
Fol' example, any grains in efficiency of tasking cannot be at the
expense of the Secretary of Defense's requirements for in-..ri7ediately
responsive intelligelice assets in crisis and in ,.lartime.
Alternative forms of resource management and operational tasking
control deserve to be.considered.
Approved For Release 2004/05/13 CIA-RDP91M00696R000200010010-3
 P a i2! 4405/13 : CIA-RDP91M00696R000200010010-3 .
l;"sciurceA tpncsi,.cqtir ~
1 . Resource decisions on collection and analysis sys te: s
and orcariizcltians could be made on a colic ial basis in a r,e o"4ia-Fing
forum for resource management matters. The Director o- Central
Intelligence or the Secretary of Defense could be.-the ch:ai rman,
but neither 'would have -Final decision authority. Differences would
be referred to the National Security Council or the President for
decision.
2. The same negotiating forum, but the DCI could have deci s1 on
authority subject. only to appeal by other members to the i;SC or the
President. -
3. The Secretary of Defense could have the decision authority..
D_ly To- , 0perat:i orlal Con, rol and Taski no, Te. chni c,!.:es
1. I fre direct tasking of intelligence colt ec--ti on syst is could
be l eft to committee decision v., i th actual comnhand authority 'refit
uncertain, and subject to neg lt"iation.
2. The President could desi mate the DCI or the Secre tart' cf
Defense as the control authority for G particular type O F sy"s to
under specific Circumstances, depending primarily oil he balance of -
importance between national and military roqui r ei ten tS.
Approved For Release. 2004/05/13 :, CIA-RDP91 M00696R00020001.0010-3-
 These v'ari ti I "1n r+ut`i1Di'1 l.f over rC=aC1LiYCC %,',, oca an and
aJc'r c1 ~1 0"lpgrc :F~i~~' L TT1"`t7YC?4 ~T/r~ d' C?I~i` 1~L~~~~I ~AAO~ReQ~010-3
r!1_1, Gwent tcechniques:
Presidential Designee
1 as ki ng Coll eg ;al i Tasking Tasking
tlanage, ,ant
Col 1 L, Cot -I'll Mlazlagl-mellt
A B
DCI Alanage:nent Control C D
Sec. Defense t anagement Control E F
Option A is about today's arranger .tints . No char es tratil d
be necessary to EO 11905.
Option B is today's resource r, inage er~t procedures wi th
more specific operational . tasking procedures . E O 11905 sign I Ci
have to be modi fi ed to, provide Specific guidelines for tasking
each generic category of "national" and "tactical" intelligence
collection.
DCI rate in resource r anagement
Option C is a s trer:gtnet;~ c~
with today's operational tasking. EO 11905 would have to be
mod i ied to provide DC.I full resource control over present BOO
assets.
Option D is a strengthened DCI role in resource ranage:.:ent
with r.oro specific operational tasking procedures. Changes in
EO 119M as in B 2 C above.
Approved For Release 2004/05/13 :?CIA-RDP91M00696R000200010010-3
 Apib3&d 9or el~ she 2B_64Th 1 3 i I$Cl9 'I WiWAK W0`2b6 Ohio='Yn
and today's by rational t:skin;. Changes would be r.ece s;:y
in EO 11905 to reduce DCZ responsibility for resource
Option F is Sec. - Der"ense resource control and' more speci.;ic
operational tasking. Changes in 0 11905 as in E Z: B above.
Possible Ojjanizational Chan jes to latch the Above t-iznaga ent Techniques
There would seem to be no organizational changes required under
al l bcati ng national intelligence resources, the question -arises whether
A and B.
With C and D where the DCI ;could assure responsibility -or
to whether the DCI as principal intelligence advisor to the President
can perform that function effectively without direct control of some
a DCI with -final authority for all resources should also remain as
head of one of the agencies (CIA) competing for these` resources-
if the answer is "no," -then the question must be addressed as
intelligence 'analysis and estimating resources. And it' he must have
such resources, what are the alternatives: The NO org ani z_:tion only,
the f10 and all of the DDI, the NO and part of the DDI?
With C Yx D, there is also the issue a i whether resource ration-
alization and efficient management would be possible i f the DCI ha d
continued to rrn.;'nage intelligence activities? This would re quire
considering the transfer of the principal DoD collection assets of
NSA, ! RO and NURP to the DCZ . Finally under C < D, -there is also
responsibility for resource allocation while the Secretary of Defense
Approved For Release. 2004/05/13 :,CIA-RDP91 M00696R00020001A01.0-3
 t~ pe to Fm,ee L the- ed of f'111 i:clco r,' nders Ci_ ;1C a ii;e o i
Approved Forelease 2004/05/13 CIA-RDP91 M0069W000200010010-3
crisis or tier' This would ieCiuiire sC,,:;:'. e>p iC'It a':is ions Tor
co'11 c gi a1 control 6f l is jar resource do ci s i at;s on roc ~;.?c: rit a,
di sconti nuance.
.In F:. Fz F the question would arise as to possible co:;' tic s
between the DCI `s role and that of SecDY i i n reso.t ce t:taraye: ent.
The DCI could be subordinated to the S;;cDef for resource. ccena e-
rrent purposes, but left with a direct access to th:: Presi da t for
purposes of advising an natters of intelligence st;bstance_
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 rh:'1'.' itro, Off )ursC, Vclr1 i? r:l oily 0.1 L11~ ,^ i E;_
Approved Foelease 2004/05/13 : CIA-RDP91 M0060002000100.10-3
especially as to d tree of control. Tilt's is, a! move to give
el the tI!a D C 1 or the SecD_., r s trolly control of
rc_sa~rce rent need not be total, but only more c'.; i rri ti ve than at present.
The attached chart SU;niilrizes these various techniques with
grada t i ons in between. .
Approved For Release 2004/05/13 : CFA-RDP91 M00696R0002000100'10-3
 Approved Fot Release 20p4/05/ ? CA-R Q
0~ X00
TIC Hr IQUES FOR MANAGING AND TAS cI G ~3 A ~1I~L N ~_?l_
DCI position
, much stronaor
.)CI position
somewhat
~~ stronger
1~ CDf_F position ,
Sole authority to dcvolop/approve.
program and budget for all
nati anal i me i 1 i cLcnce ar.i i Vi ti es
Sole authority for operational' I
tasking of all national i ntel -
l i pence ecti.yi ti es
---Tpro~C ram/buudget controls
Operational tasking
--j
t--- Program/budget control s
DCI not operating head of CIA;
remove NSA and National Recon
naissance from DoD; CIA, NSA
and National Reconnaissance under
direct DCI operating control
E.G. 11905 amended to clarify DCi
.
authority to deal directly with pro(
managers and heads of intelligence
offices
E.O. 1*195 amended to strengthen DCI
authority to task NSA and National
Reconnaissance el ements directly
SCI and NSC/PRC position on NFIP budgetl
to be a recommendation to SECDEF in
novel oFjment of 'his Departmr: h rdoot
Iscme;what
Istronrer
1SECDEF position
tmuch stronger
No DCI tasking of DoD intelligence
Gi(r ~erati oval as ki nn i-e? e :ones excep L hrough, and with
ar)proval of, SECDEF
-
Sole authority to develop/approve
program and budget or all h
national iPL?ol1igcnce activities
l i cienco activities
Sole authority for operational
-J, tas ki na of all ?. nati oval Intel-
IDCI heads CIA but reports to SECDEF fort
~' resource management purposes
Approved For Release 2004105/13 : CIA-RDP91 M00696R000200010010-3
-._.,Re tai i n present situation No change
 25X1 Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
Next 1 Page(s) In Document Exempt
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 Approved For-Release 2004/05/13 : CIA-RDP91 M0069MR000200010010-3
DRAFT
20 Apr 77
M. S.
1. The discussion yesterday on organizational issues settled, it
seemed to me, around two questions:
a. While it is desirable to merge the DDI and the NIOs
and retain both in CIA, how can CIA produce national
intelligence immune from contamination by CIA's organizational
biases?
b. The NI0's multiple roles distract from the estimative
function, but, if the NIOs do not fill these other roles, who will?
2. Addressing the first issue, one possibility would be to
distance CIA's production offices from its collection offices by having
the production offices report to the DCI through principal deput s,
the D/DCI/NI, who would wear two hats like the DCI. Certain of these
offices would product national intelligence (current and estimative) and
the rest would do research in support of the national products; the
distinction would be whether the product were coordinated or not. This
would, essentially, take us back to where we were organizationally in 1965.
3. 1W the solution to the first issue would leave the D/DCI/NI
in charge of a variety of overlapping, functionally focussed offices--
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 Approved For elease 2004/05/13 : CIA-RDP91 M006 000200010010-3
estimative, economic, current, strategic, political, scientific, etc.--
the management of which would potentially be a nightmare. Which brings
us to the second issue. For an answer, I would look to the analogy of an
insurance conglomerate which finds itself in a variety of businesses--
fire, casualty, life, reinsurance, finance, banking, etc.--but organized
into a number of companies which overlap in the coverage they offer, sing
in the clienteles they serve, and in the geographic regions they work.
The way an insurance conglomerate solves its managerial problem is to
have functionally delineated vice presidents who have line responsibility
over all the companies in the conglomerate. I suggest, in other words,
that a D/DCI/NI, served by a series of intelligence officers with line
responsibilities delineated much like those of the present NIOs, would
be able to manage the monster at his disposal. These officers, supported
by small staffs, would fill the NIO's customer relations, representational,
and advisory roles, plus be the D/DCI/NI's troubl e-shooters -moni to/of , the
quality of the prodcuts.
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 D Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010Qj 0-3
R `.' R
A A
F F
T T
1. The dilemma facing us is an old familiar one. Given the
President's commitment to rationalization, there are only three out-
comes to PRM 11 that have much chance of serious consideration:
-- Deferral of major changes to give a further
test to E.O. 11905, along with some minor modification
thereof.
-- Increase in the DCI's budgetary authority
in ways that make him and SecDef partners in
managing the intelligence business.
-- Assignment to the DCI of line authority
over NSA and NRO as well as CIA.
2. The question that arises in all three variants is the DCI's
relationship to CIA. DoD would like to see the first option, with the
DCI separated from CIA as well. This is unlikely, but for the second
and third options the DCI must ask whether he should balance the increase
in his authority elsewhere by a decrease in that over CIA. One has
to assume that the D/CIA would in any case report to the NSC through
the DCI; any other arrangement would be asking for trouble. On the other
hand, if he does report through the DCI, what difference have you made?
Perhaps the best way to think of the resultant arrangement would be the
present relationship of SecDef to NSA: He controls the money, and he
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 Approved her Release 2004/05/13: CIA-RDP91 M06R000200010010-3
4
-probably knows what Lew Allen looks like, but he hasn't ever been there.
3. Should something like this come to pass, the future of CIA
is called into question. If the DCI is separated in the manner adum-
brated above, should he take the DDI along as well as the NIOs? Or
can he get along with the NIOs as they now are? Or augmented? An inter-
locking question: ire move out of Langley completely, where should
his substantive staff, however constituted, be? And is this feasible?
Would an NIO staff, supporting a DCI separate from CIA, be able to give
sufficient attention to the estimative function? Or would it be necessary
to create some separate organization, insulated from fire-fighting, to
do this? Is an NIO/DDI merge in a national production organization
under the DCI a good idea or a bad one?
Approved For Release 2004/05/13 : CIA-RDP91 M00696R000200010010-3
 ApprovIdf0ft
3nC8-k M 00696R000200010010-3
SUBJECT Organizing for Estimates
I will have to miss the first half hour of
tomorrow's meeting on this subject, so I took
the opportunity to set down a few thoughts on
what seemed to me fixed desiderata in almost
any scheme for reorganization of estimates.
NIO WE
Approv
Date 18 April 1977
 Approved For' elease 2004/05/13 : DP91 M0069WO00200010010-3
~y a
S E
18 April 1977
Functional Propositions re Organizing National Intelligence
1. The answer to almost any question about organizing for national
intelligence depends on the answers to other questions.
-- Thus the answers to where to locate or whether to merge
NIOs and the DDI depends greatly on how much one wants to
make national intelligence a community product, versus a
CIA product, or how much long-range analysis the DCI wants
to supply as opposed to short-order policy support on tac-
tical questions. And of course the answers are usually
that he wants both.
2. Since these interdependent questions thus lead to a circular
exercise, it seems more fruitful at this stage to set forth a few propo-
sitions which (to me) appear valid almost regardless of organizational
structure. These derive from the performance of the NIO business as I
have seen it. They are not meant to be as critical of the NIO system as
they may sound, nor as uncritical of ONE as they appear by implication.
It is simply that we are approaching decisions that, however they go,
should and can be made in the light of experience rather than theory.
A. NIOs have large theoretical responsibilities and no
commensurate assets.
-- They have contrived to get estimates and other analysis
done because successive Directors (especially Colby)
frequently made clear that they expected the system to
work. And on some problems it has worked pretty well.
These problems are either the kind where the same
offices and people are doing the job that always did
the job (i.e., Soviet military) or the problem was
relatively simple (i.e., one country or region, and
one-dimensional -- political or military or economic).
Where it crosses borders or disciplines, we have been
less successful. Where it has been a question of
challenging long-range conventional wisdom or exploring
new frontiers, we have been less successful than on
tactical, short-term needs. And where time is urgent,
we have often experienced chaos, since when subjects
25X1
Approved For Release 2004/09
6-0 ?ELFIDP91 M00696R60200010010-3
 S Er)'
Approved For`I(elease 2004/05/13: 6A-RDP91 M0069W000200010010-3
-2--
heat up, that is precisely when it is hardest to get
the best analytical talent to work on estimates in the
face of competing demands by home offices, the source
of the incentives.
B. NIOs are spread too thin.
-- If estimates are to be taken seriously, they must be
done by people with the time and inclination to take
them seriously -- not as afterthoughts or side-interests
on the part of people heavily engaged with customer
relations, quasi-managerial tasks in support of the IC
Staff, current briefings, and other functions.. This is
true in spades if, as is often the case, the NIO has
personally to do a lot of the drafting, most of the
chairing, and much of the other staff work on estimates.
C. Estimating is a discipline.
-- Estimating is analysis with a forward reach. Like good
analytical history, it is an art requiring a certain
cast of mind and training. In addition to the universal
virtues of objectivity, and clarity of thought and
writing, estimating for top policymakers requires unusual
capabilities for selectivity, ability to boil down masses
of information and to generalize non-superficially, to
keep topical and relevant without being captured by cur-
rent headlines and intellectual fashions. It even
requires a degree of imagination and savvy, as to what
questions are important to customers, in order to avoid
writing academic treatises for other specialists.
D. Thoughtful estimating requires the right conditions.
-- If the menu is to live up to the advertising, one must
get good chefs, provide the right ingredients and kitchen,
and then make sure that they are permitted to cook and
are not also kept busy as waiters, busboys, and enter-
tainers. It costs some money to do this, but one cuts
corners only at considerable cost in quality and ulti-
mately in reputation. The NIO system has cut many cor-
ners in the name of economy, anti-elitism, getting closer
to the policy process, etc. Some corners have been cut
harmlessly, perhaps a few even profitably. But if the
whole restaurant is to be reorganized, it is time to
make sure conditions and people in the estimates depart-
ment of the kitchen are such that performance can live
up to promises on the bill of fare.
Approved For Release 2004/05/1q ff91 M00696R000200010010-3
0 L u.u f 'i.._
 I;~ o
Approved Forelease 2004/0/'h3`:' K` DP91 M006000200010010-3
- 3 -
E. All the above implies that --
-- If an estimates staff is to be established, the
crucial things are
-- that it be manned by the right people,
specifically a staff of 15 to 20 first-
class analytical/estimative drafters,
-- that the staff be chosen by and responsive
to the entity charged with producing esti-
mates (whether present NIOs, a collegial
review authority, or whatever),
-- that it be sufficiently insulated from cur-
rent support and other functions to be able
to sustain quality performance and produce
in timely fashion. (It must also, of course,
be sufficiently "wired in" so as not to
become another ivory tower like the late
OPR.) The', need not estimate eight hours a
day, but i:. is critical that they consider
it their primary job, and
-- that grade structure, physical environment,
etc., must be such as to attract the cream of
the crop from within the Agency, the community,
and outside.
F. Creation of such a staff would inevitably bring some
criticisms about elitism, threats to the rice bowls of existing.-
shops, etc. None of these would be unmanageable or even very
serious if the DCI simply made clear that he wanted it this way.
Cooperative working arrangements with existing offices would be
less difficult to work out than current arrangements have been.
And in many quarters of the DDI, DIA, and INR the net reaction
would be one of relief.
Approved For Release 2004/05:Cr
P91 M00696R000200010010-3