DCI'S ROLE IN WARNING AND CRISIS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP81B00493R000100050001-7
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 28
• Document Creation Date: December 15, 2016
• Document Release Date: July 6, 2004
• Sequence Number: 1
• Publication Date: September 8, 1978
• Content Type: MF

SECRET
Approved FRelease 2004/08/16: CIA-RDP81B0046R000100050001-7
8 September 1978
MEMORANDUM FOR: Director of Central Intelligence
THROUGH: Deputy Director of Central Intelligence
FROM: Richard Lehman
Associate Director for Substantive Support
SUBJECT: DCI's Role in Warning and Crisis
1. The attached memorandum signed byl and Sayre Stevens
for Bob Bowie responds to your comments on the working group's study of
your role in crisis and warning. It contains, as you requested, a broad
range of options on which we ask that you indicate your tentative pre-
ferences, but makes no recommendations.
2. I have also attached a matrix and a decision tree which are
designed -- we hope -- to help you find your way through these complex
matters.
3? believes that any of the three options for lodging
the leadership function can be made to work satisfactorily and that
regardless of where it is lodged the function should take the form of
a responsible individual whether his title be Strategic Warning Officer,
NIO for Strategic Warning, NITO for Crisis Management, or whatever.
However, he recommends that a decision with respect to the disposition
of the Strategic Warning Staff should not be made in. isolation from
other issues concerning movement of collection tasking organizations to
the Pentagon.
4. We have not been able to consult Bob Bowie. Sayre and I think
he would vote for a variant of Model B (B1 on the charts) which placed
overall responsibility in NFAC, but replaced the "center" with a single
staff officer, perhaps an NIO.for Warning. In general, he believes
elaborate systems are unnecessary.
25X1
25X1
Approved For Release 2004/08/16: CIA-RDP81 BOO
 SECRET
Approved Farr#2elease 2004/08/16: CIA-RDP81B00R000100050001-7
5. Two significant dates are approaching. The EPSCI Subcommittee
on Evaluations has tentatively scheduled hearings for 19-20 September on
its report on warning. (There may be some give to these dates). More
important, Frank Carlucci is scheduled to be the keynote speaker at the
DIA I&W Seminar on 27 September.
6. Mr. Bowie and I have noted to me that they look
forward to discussing these issues with you.
25X1
Distribution:
1 - DCI
1 - DDCI
1 - ER
1 - DD/NFA
1 - DD/CT
1 -
1 -
1 - AD AC ss
1_- AD/NFAC/sS Chrono
NFAC Registry
1 - Warning & Crisis Management File
Approved For Release 2004/0?M :E(CIA-RDP81B00493R000100050001-7
 Approved FO&Release 2004/08/1 6 B CIATRDP81 B004R000100050001-7
8 September 1978
MEMORANDUM FOR: Director of Central Intelligence
THROUGH Deputy Director of Central Intelligence
Robert R. Bowie
Deputy Director for National Foreign
Assessment
Deputy Director for Collection Tasking
SUBJECT The DCI's Role in Warning and Crisis
REFERENCES : a. Report on the DCI's Role in Warning
and Crisis, dtd 22 June 1978
b. Memorandum from DCI to DDCI, same
subject, dtd 18 July 1978
Introduction
1. Your memorandum of 18 July notes that, although
all the relevant sources and experiences are cited in the
Working Group's report, the report does not lay out "a
series of alternatives between which we can exercise a
decisionmaker's judgment." You asked for a statement of
the essential elements of warning, some alternative ways
to achieve an adequate warning program, and an evaluation
of previous organizational arrangements for warning.
2. Knowing that you have read the report, we have not
attempted to rewrite it or to review again for you the argu-
mentation and background. Rather, this memorandum is con-
fined to the questions you raised.
3. Our approach to providing you the clear alterna-
tives you ask for is necessarily somewhat complex, reflect-
ing the intractability and intertwining of the issues.
Section I of this discussion lays out the minimum require-
ments for a national warning system as a yardstick against
which to judge alternative approaches. Section II develops
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
SECRET
 SECRET
Approved FbooRelease 2004/08/16 : CIA-RDP81B00406R000100050001-7
a number of models of national warning systems, past,
present, and possible, and measures them against the yard-
stick of Section I. It will be apparent that in deciding
among these models, or considering others not discussed,
you will have to make certain fundamental choices. These
are presented in Section III.
4. The Working Group noted that warning and crisis
management probably should be managed together, as related
functions involving many of the same people and organiza-
tions, but that substantive operations should be kept sepa-
rate, in order that a crisis in progress not obscure the
potential emergence of another. For that reason, we ad-
dress crisis management in Section II by noting when par-
ticular arrangements for Community coordination of warning
policy and procedures are not suitable for a parallel co-
ordination of crisis management matters. For the substan-
tive side of crisis management, the Working Group did not
recommend any change from the approaches you had already
evolved ad hoc. We do not, therefore, review these here,
but a few issues need to be settled. These are included
in Section III.
5. When you are ready to express your tentative pref-
erences on the issues here presented, we recommend you meet
with us for a review before finally committing yourself. At
that time we should also discuss how best to engage the rest
of the Community.
I. Requirements for a National Warning System
6. The discussions in the Intelligence Community over
the past few years indicate a consensus as to the essential
ingredients of a warning program. The wording varies from
one forum to another and from one warning and crisis study
to another, but one can perceive these essential principles:
-- Warning must be an explicit mission of all
intelligence organizations.
-- There must be a way to converge intelligence
information in order to analyze it for I&W
content.
-2-
Approved For Release 2004/08/16 CI RDP81B00493R000100050001-7
 SECRET
Approved F i2elease 2004/08/16 : CIA-RDP81BO0 R000100050001-7
-- The output must be recognizable as warning.
-- The output must flow up, laterally, and down.
7. The complexity of the warning mission has in-
creased rapidly in recent years. The indications of im-
pending crises come to us from a wide variety of geograph-
ical and functional specialties within and without the
Intelligence Community. We must assign warning as a mis-
sion for all intelligence organizations because of the many
sources of indications and in the interests of economy.
We cannot afford to duplicate the Intelligence Community
with an apparatus devoted solely to I&W.
8. There is a requirement that intelligence informa-
tion converge in order to analyze it for warning. This
implies both a technical capability for handling informa-
tion and a means of focusing organizational activity.
9. The intelligence message must be clearly recog-
nizable as a warning. The last thirty years are littered
with crises where the indications were perceived, evaluated,
and passed on to military operators and national decision-
makers but the warning message was not effectively communi-
cated. There were several reasons for such failures. In
some cases the intelligence analyst simply failed to recog-
nize the indications of a crisis. In many more cases, how-
ever, the message lacked a warning label because the sender
did not have an explicit warning responsibility and a con-
comitant authority to send a "warning." In other cases,
the military operator or national decisionmaker failed to
heed the warning because the sender was not "the official
warning office."
10. There must be an established and readily recog-
nizable means whereby the output of a national warning
system flows up to the President, laterally to other de-
partments, and down to the military operators. Placing
all on the same footing is an obvious requirement. It
is less obvious that the national warning system is de-
pendent upon inputs from the same sources to which it owes
warning.
-3-
SECRET
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
 SECRET
Approved FbirRelease 2004/08/16 : CIA-RDP81 B00 R000100050001-7
11. Just as there is general consensus on the essen-
tials of a warning system, there appears to be consensus on
the functions that such a system must incorporate. These
are:
-- Coordination, across the Community and
across disciplines, of warning management,
policy, procedures, and methodologies.
-- Analysis, the identification, convergence,
and assessment of warning information and
the formulation of the warning message.
-- Discipline, the means by which Community
line organizations, which have primary re-
sponsibilities other than warning, are kept
sensitive to their warning responsibilities.
Challenge, the insurance taken out against
analytic failure.
II. Systems for Warning
12. There are any number of ways in which these func-
tions can be wired together in a national system. The
range is bounded at one extreme by an integrated collec-
tion and analysis system fully dedicated to warning and
at the other by no system at all. The first is unaccept-
able as enormously expensive and duplicative (warning is
an integral part of all analysis), the second is equally
unacceptable both politically (the DCI's "Pearl Harbor"
responsibility) and practically (central coordination is
needed).
13. In this section we analyze both systems used in
the past, as you requested, and some other approaches,
against the set of requirements postulated in Section I.
14. The "Watch Committee" system used from 1951 to
1974 consisted of a senior interagency committee, usually
chaired by the DDCI, that prepared weekly and occasional
special Watch Reports, and a 24-hour jointly manned Na-
tional Indications Center under a CIA Director. The
Approved For Release 2004/08/16'E61fkDP81 B00493R000100050001-7
 SECRET
Approved Foa.Release 2004/08/16 : CIA-RDP81B0044MR000100050001-7
Committee submitted its reports to USIB (now NFIB) and
the DCI issued them after USIB discussion as is still
done with Estimates. The NIC was linked to the working
levels in each intelligence agency through a network of
staff officers who acted as warning referents.
15. This arrangement provided all the functions
needed for a system. The Watch Committee, assisted by
the NIC and staff arrangements throughout the Community,
handled both the coordination of policy and operations
and the analytic function; the DCI had a central focus
for strategic warning. (Strategic warning was vaguely
defined, but the overwhelming emphasis was on Soviet
military attack.) The Community-wide review that was
required. to prepare for the weekly Committee meetings
provided the discipline necessary. The Committee and the
NIC, in theory at least, performed the challenge func-
tion from a position partially independent of the current
intelligence apparatus.
16. This system eventually failed, not because it was
ill-conceived, but because the world in which it functioned
changed and it did not.
The intense national concern with surprise
Soviet attack which had caused its creation
gradually decreased. This led Community man-
agers to give a lower priority to the assign-
ment of good people to the warning apparatus.
In time, the NIC became a turkey farm.
As a consequence, the NIC was no longer capa-
ble of performing a challenge function and
could not command the respect of line organi-
zations in this role.
The Watch Committee/NIC missions became con-
fused with the current intelligence mission.
The result was both to diffuse the warning ef-
fort and to place it in competition with cur-
rent intelligence.
-5-
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
SECRET
 Approved Fo Release 2004/08/16: CS IA RDP81B00 8000100050001-7
-- The intelligence business matured mightily in
the two decades after 1951. Whereas the NIC
was the only thing of its kind at the begin-
ning of the period, it had been badly over-
taken by the agency operations centers at the
end of it in terms of facilities, communica-
tions, and access. If it was to play a role
in the 1970s, it would need an extremely ex-
pensive modernization that would unnecessarily
duplicate existing facilities. It was, in
effect, a fossil of the 1950s.
.
-- Handling substance through weekly committee
meetings both at the Watch Committee and the
USIB (NFIB) level, was barely workable in the
1950s, but was never the most effective way
of developing a clear warning message. By
the 1970s, it had degenerated into haggling
over the wording of current intelligence and
was clearly ineffective.
17. The reorganization of 1974 abolished the Watch
Committee and the NIC. The Deputy Director for Production,
DIA, was designated by the DCI as his Special Assistant for
Strategic Warning. Under him, a jointly-staffed Strategic
Warning Staff was established with a CIA Director. "Stra-
tegic Warning" was more precisely defined to deal only with
military attack by Communist powers on the US and its al-
lies. (This is the definition referred to below as the
"narrow" warning mission.) The SWS was to be a challenge
mechanism and to conduct research in the indications and
warning field.
The Special Assistant, with the advice of the Director,
SWS, was to be responsible for recommending to the DCI
the issuance of Strategic Warning Notices. (None has
ever been issued.) The strong points of this arrangement
were that it gave full recognition to the DoD role in
strategic warning and that it provided a direct and un-
cluttered channel for the warning message. But if the
Watch Committee system had been ponderously bureaucratic
and stultified, its replacement went too far to the other
Approved For Release 2004/08/1 6ERDP81 B00493R000100050001-7
 SECRET
Approved FocoRelease 2004/08/16: CIA-RDP81B0049OR000100050001-7
extreme. In eliminating outmoded organizations and medi-
ocre personnel, we also eliminated the critical underpin-
nings of a national. warning system. The terms of the es-
sential functions noted above:
-- The Special Assistant--a senior DIA officer--
was supposed to combine, under the DCI, the
coordination of policy and operations and the
analytic mission. In fact, in his anomalous
position--with secondary duties assigned by
an authority whose interests often are dif-
ferent from those of the authority for whom
he performs his primary ones--he could not
play a central coordinating role. The ar-
rangement was widely seen as a retreat by a
besieged DCI (Colby) from his warning respon-
sibilities. The Special Assistant has exer-
cised his authority only within the DoD chain
of command.
-- The SWS has no formal links to the rest of
the Community and there is no community-
wide warning organization or routine.
Warning outside DoD is entirely a current
reporting responsibility. Thus the SWS is
unable to serve as an energizing force for
warning matters; the warning discipline
that might sensitize the Community is lacking.
-- Without structured links to the Community the
SWS is ineffective in its challenge role. Many
analysts are unaware of its existence. More-
over, although it did not inherit the personnel
or practices of the NIC, it suffers from the
same manning problems.
-- The narrowed scope of the strategic warning
mission omits a wide range of the warning
spectrum. (Referred to below as the "broad"
warning mission.)
Approved For Release 2004/08/161$P81 B00493R000100050001-7
 SECRET
Approved Foo.Release 2004/08/16: CIA-RDP81B00499R000100050001-7
19. New Models. Current pressures for structure and
order, e.g., from HPSCI, are a recognition of these weak-
nesses. Some relatively feasible and inexpensive ways of
repairing them are as follows:
A. Fix the present system. This would re-
quire stronger links between the Special Assistant
and the DCI on the one hand and the SWS and Com-
munity analysts on the other. It could be done
by:
-Creating a DCI Committee on Warning
chaired by the Special Assistant, and
charged with the coordination of warn-
ing policy and operations for the
Community (but not the analytic
mission).
-- Reaffirming and publicizing the special
Assistant's responsibility to the DCI
for warning analysis.
-- Broadening the strategic warning mis-
sion to include warning of any situa-
tion that might lead to US-Soviet con-
frontation. (A compromise between
the narrow and broad missions.)
-- Reaffirming the challenge mission of
the SWS and requiring the Community
to upgrade its personnel.
-- Providing discipline by charging the
SWS with conducting a weekly review
to sensitize the Community to warning
matters; designating referents in
each agency through which the full
analytic resources of the Community
participate in these reviews.
Strengths
-- Simple and inexpensive.
-- Least disruption.of present arrangements.
-- Recognized DoD role in strategic warning.
-8-
Approved For Release 2004/08/1 ~E8 kDP81 B00493R000100050001-7
 SECRET
Approved FbwRelease 2004/08/16 : CIA-RDP81B00 R000100050001-7
Weaknesses
-- DCI is carrying out his most important
single responsibility through an officer
not subordinate to him.
-- Focus of coordinating committee in DoD
incompatible with DCI's crisis manage-
..ment responsibilities to President and
NSC.
-"The more mission is broadened to include
warning of the politico-military and
economic events that are realistically
most likely to matter to the US in the
next few years, the less appropriate
lodging this mission in a DIA office
becomes and the less capable of deal-
ing with them is a small SWS.
-- Conversely, the more narrowly the mis-
sion is defined, the larger that slice
of the spectrum not covered by any struc-
tural warning system.
-- A SWS charged with warning of events
that are intrinsically unlikely will
issue warning very rarely. Either,
like the NIC, it will atrophy and fail
to warn when it should, or it will go
looking for another mission and confuse
Community functions and chains of com-
mand.
-- Manning the SWS will be no easier than
before.
-- There is no clear role for NFAC, the DCI's
own analytic organization, in the warn-
ing chain.
-- The responsibilities of the Special
Assistant may overlap or conflict
with those of DD/CT and DD/NFA.
Approved For Release 2004/08/16 : CI"W$00493R000100050001-7
 Approved F .Release 2004/08/16 : CIA- I 00 R000100050001-7
B. An NFAC Warning Center. This model would
emphasize the importance of the analytic process in
warning.
-- Scrap the present system.
-- Designate a senior NFAC officer as the
DCI's warning and crisis management
officer.
--.Make him chairman of a DCI Committee
charged with coordination of policy
and operations.
-- Create under him a "Warning Center" in
NFAC staffed with perhaps 25 profes-
sionals drawn largely from NFAC but with
at least some Community participants.
It would incorporate, but not control,
the NITO for Warning and other NITOs as
appropriate. The Center would be re-
sponsible for all analytic aspects of
warning under a broad definition.
-- Provide a Community-wide discipline by
requiring the rest of NFAC and other
Community agencies to conduct regular
warning reviews and provide the results
to the Center.
-- Encourage challenge and debate among the
Center, NIOs, and line organizations.
Strengths
-- Recognizes importance of analytic process
in warning.
-- Provides lively challenge function.
-- Center large enough and stimulating
enough to avoid stultification, will
be much easier to man.
-- Center fully able to absorb and analyze
warning information.
Approved For Release 2004/08/16: CIASEUR I 00493R000100050001-7
 SECRET
Approved Foot elease 2004/08/16 : CIA-RDP81B004 2000100050001-7
Provides a clear focus for the warn-
ing information flow Community-wide.
Weaknesses
-- Expensive.
-- Minimizes attention to management and
collection aspects, especially in crisis
management.
-- May overlap with responsibilities of DD/CT.
Community equities given short shrift.
Overlap of analytic missions will lead to
unnecessary bureaucratic frictions, even
paralysis.
-- Danger of consumer receiving contradic-
tory interpretations because two analytic
channels exist.
-- Community participation likely to be
pro forma.
C. A Community Coordination Facility
-- Scrap the present system.
?- Place responsibility for coordination
of warning and crisis management policy
and operations on the DD/CT with staff
responsibility assigned to the NITO for
Warning and Crisis Management based in
the Pentagon.
-- Place responsibility for analytic and
,production aspects on the DD/NFA (NFAC),
and establish an NIO for warning on his
Staff.
-- Define warning broadly.
Approved For Release 2004/08/16 : Cl1 'B00493R000100050001-7
 SECRET
Approved FdRelease 2004/08/16: CIA-RDP81B00499R000100050001-7
Assign the challenge function jointly to
the NIO and NITO for warning, each ap-
proaching the problem from the perspec-
tive of his particular discipline.
-- Provide a network of warning referents
in each agency through which the NIO and
NITO for Warning can sensitize the Com-
munity in their respective areas of
cognizance.
Strengths
- Consistent with DCI's reorganization of
functions among his deputies.
-- Gives major attention to collection and
interagency coordination aspects.
-- Basing coordination element (DD/CI) at
Pentagon accommodates DoD equities;
facilitates DCI "shift of flag" to main-
tain leadership in military-related
crises.
-- Meets HPSC(I) strictures to integrate
and use existing DoD capabilities.
-- Least expensive.
Weaknesses
-- Warning system focus is split organiza-
tionally (HPSC(I) called for a single
point of focus) between CT and NFA and
in location between the Pentagon and
Langley; split could cause disconnects,
e.g., between collection and production.
-- Community equities not fully recognized
on the analytic side.
-- NIO for Warning may not have bureaucratic
position to perform effectively, particu-
larly if warning is broadly defined.
Approved For Release 2004/08/16 : CIA- r 00493R000100050001-7
 SECRET
Approved FoK&elease 2004/08/16 : CIA-RDP81BO049 8000100050001-7
D. A Hybrid System. This is the solution recom-
mended in the Working Group study. It attempts to
reconcile the diffuse and unpredictable requirements
of warning as broadly defined with the critical and
often highly specialized requirements of strategic
warning as narrowly defined. It would:
-- Scrap the present system (but see below).
-- Create a unified system with the DCI or
DDCI at the apex, acting through a senior
Warning (staff) Officer who would also be
responsible for crisis management. (The
SWO might be a DoD officer on detail to
the Office of the DCI.)
Retain the SWS under the SWO, but improve
quality of manning.
-- Coordinate Community policy and operations
through a senior interagency steering
group chaired by DDCI or a DCI Committee
chaired by the SWO.
-- Handle analytic aspects of warning (broadly
defined) through the NIOs acting for the
Community, with the SWO having an addi-
tional challenge responsibility ("ombuds-
man for warning"). Provide a Community
discipline by requiring each NIO to con-
vene analysts periodically for discussion
of future contingencies; each NIO to re-
port results to SWO.
-- Handle analytic aspects as narrowly de-
fined through the SWS in consultation
with certain NIOs. Provide a Community
discipline through reestablishment of
referent network, with periodic stra-
tegic warning reviews provided to SWS.
SWS retains its present challenge
function.
-13-
Approved For Release 2004/08/1 %ERDP81 B00493R000100050001-7
 SECRET
Approved ForIease 2004/08/16 : CIA-RDP81B0049.&R000100050001-7
S trencths
-- Demonstrates importance DCI gives to criti-
cal mission.
-- Places unified responsibility at DDCI
level where collection and production
threads come together, keeping manage-
.ment lines clear.
-- Recognizes most Community equities.
-- Places primary challenge function on
officers best equipped to perform it
(NIOs).
-- Encourages through NIOs broadest Community
sensitization to warning matters.
-- Accommodates both broad and narrow warn-
ing functions.
Weaknesses
-- More complicated than other models.
-- Would require a few more people than
present system.
-- Uncertain whether SWS can be made effec-
tive.
May not go far enough toward Community.
E. A Variant to D. You have expressed some skep-
ticism as to t e use u ness of SWS. It would be pos-
sible to eliminate it from Model D, but it would remove
a key element from the system. The SWS is the instru-
ment by which Community discipline with regard to the
narrow warning mission would be enforced; and an im-
portant node at which strategic warning information
would be converged. It could be replaced by:
-- Placing an additional burden on the
NIO/SP, NIO/CF1 and NIO/USSR-EE;
-14-T
Approved For Release 2004/08/16ROA-RDP81 B00493R000100050001-7
 SECRET
Approved Fo lease 2004/08/16: CIA-RDP81BO049? 2000100050001-7
Creating an NIO for Strategic Warning; or
Providing the SWO with a staff.
There are difficulties in all these, but the NIO/SW
is the most attractive.
As compared with Model D, Model E with the NIO/SW
would:
-- Be less expensive, as SWS positions could
,be used to provide the staff positions
called for in the agencies and under the
DCI.
Provide somewhat less attention to the
narrow strategic warning mission.
Have slightly less "Community" flavor.
Be somewhat simpler in structure and
function.
III. ISSUES
20. In the preceding section five broad organizational
structures were discussed; each represents a different com-
promise among the policy issues that await early decision.
In the following sections each of these policy issues is
discussed in its own right. They are, however, interrelated
and cannot be decided in isolation.
Warning Issues
21. The HPSCI recommends, "That the DCI provides a
focus for warning leadership in the community, which may
require appointment of a special assistant for warning."
Most observers agree there is such a need, but not all
agree about where to lodge the focus function and about
its forms. How these are decided will largely determine
the balance struck between two sets of competing demands.
The first set is how to balance the demands that arise
from military requirements against those that arise from
-15-
Approved For Release 2004/08/1 e -RDP81B00493R000100050001-7
 SECRET
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
the broader needs of the President and his foreign policy
advisers. The second set is how to balance the attention
given to analysis against that given to collection. Cur-
rent analytic weaknesses are of more immediate concern
from a warning viewpoint. But, collection involves far
more resources; errors there can waste a lot of money.
Also, tasking collectors in a crisis will present some
contentious decisions.
22. Leadership can be placed in:
-- O/DCI (Models D and E)
Pro -- Symbolic of importance of warning.
-- Ready access to DCI.
--.Clearly a Community position.
Con DCI spread too thin to give it attention.
-- Could be perceived as subverting chain
of command.
-- O/DDCI (Models D and E)
Pro -- Symbolic importance still there.
-- Moderate access to DCI.
-- Lowest point at which collection
and analysis chains of command come
together.
Con -- Could be perceived as subverting chain
of command.
-- O/DD/NFA (Model B)
Pro -- Warning is largely an analytic problem
and here's where the assets are.
Con -- Tendency for Defense to see this as a
"CIA" position.
-16-
SECRET
Approved For Release 2004/08/16 : CIA-RDP81 B00493R000100050001-7
 SECRET
Approved Fo eIease 2004/08/16 : CIA-RDP81B0049aR000100050001-7
Relatively reduced access to collectors,
especially for planning crisis management
preparations.
-- In-house coordination a problem.
-- O/DD/CT
Pro -- Seen by Defense as more "Community"
than DD/NFA.
Here is where control of collection
assets used in crisis management will
be exercised.
Con -- Reduced access to analysts who are pri-
mary warning source.
-- In-house coordination a problem.
O/DD/RM
Pro -- Short of DCI the Deputy most seen by
observers from outside CIA as "Com-
munity."
-- Has expertise to examine programs and
evaluate efforts.
Con Has direct control of no analytic or
collection tasking assets.
-- Assignment of this function would re-
quire broadening charter somewhat.
Defense Department (Model A)
Pro -- Access to elaborate DoD I&W mechanisms.
Con -- Split responsibility to DCI and SecDef.
-- Lack of access to DCI and political
analysis.
-- Will not be seen as Community.
Approved For Release 2004/08/16 : CIA'TB00493R000100050001-7
 SECRET
Approved Fo telease 2004/08/16 : CIA-RDP81B004QaR000100050001-7
Or it can be split. (Model C) One possibility
is to split responsibility between DD/NFA and
DD/CT. The factors arguing for or against any
such split are:
Pro -- Warning and crisis management cross
existing organizational responsibili-
ties. So putting focal point under
any one Deputy (except DDCI) will do
violence to the existing chain of
command.
Con -- Splitting the function will be seen by
most observers as reducing the impor-
tance attached to fulfilling the warning
function.
23. A secondary question is what form the leadership
should take. Should it be lodged in an individual (and per-
haps an assistant), an interagency committee, or an operat-
ing organization of up to 25 people? This turns out not to
be a choice, however. Any individual with staff responsi-
bility under the DCI for warning is going to need some sort
of interagency committee or working group through which to
coordinate Community activities. And a committee is going
to need a chairman. (It is generally agreed that such a
committee should be managerial and should not be involved
in the substantive process of warning.) Further, any
operating organization will have to be headed by an indi-
vidual who serves as the DCI's "leader" or answers to the
leader. The question then is really whether the individual
needs to be backed up by such an organization, as by a re-
juvenated SWS in Models A and D, or by an NFAC center in
Model B.
Pro -- Such a staff can do independent warn-
ing analysis.
-- Can perform an effective devil's advo-
cacy role in the community.
-- Provides locus of responsibility for
producing the warning message.
Approved For Release 2004/08/16SFiTDP81 B00493R000100050001-7
 Approved For Release 2004/08/165 C RDP81 B004R000100050001-7
Con -- Costly.
-- Problem keeping such a group relevant.
They tend to become isolated and mori-
bund.
-- Competition with line organizations.
24. A cross-cutting question is the breadth of the
Community warning effort. Should it deal with all major
developments of concern to national policymakers (Models
B and C), only with military attacks on the US or its al-
lies (present system), or something in between (Model A)?
One can also visualize a hybrid system in which broad warn-
ing issues are treated one way and narrow issues another
(Models D and E). Such a hybrid is inevitable as a practi-
cal matter if there is to be a class of warning problems
on which special techniques and talent are to be brought
to bear, because there will always be some warning problems
that lie beyond the pale defined by resource limitations.
Therefore, the question reduces to, "What warning problems
should be subjected to special formal warning procedures
and who should operate them?"
25. One can subject one or more of the following
warning problems to special procedures.
-- Military attacks on the US or its allies by the...
...Warsaw Pact ...PRC
...North Korea ...Others?
-- Situations that might lead to a US confronta-
tions with...
...USSR ...Others?
Which of these or other problems to subject to special pro-
cedures involves balancing the extra assurance of producing
a warning which is implicit in using more formal procedures
on and devoting more resources to these problems against the
SECRET
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
 SECRET
Approved Fos.Release 2004/08/16 : CIA-RDP81 B0049GR000100050001-7
overall cost and effectiveness of the Community warning
system and against the fact that most of the "likely" (not
necessarily most dangerous) events of which national policy-
makers would want warning are too diffuse and numerous for
application of a highly structured approach.
26. Having defined the limits above one then asks,
"How to do it?" There are several options.
-- Make Defense the DCI's executive agent for
narrowly defined warning tasks.
Pro -- Most of the structured formal warning
procedures are already under DoD aus-
pices and the main impetus for such
work comes from military commanders.
Con -- This is unlikely to work except for
warning of war and in any case will
not be seen as a Community function.
The man in charge will inevitably be
torn between two loyalties.
-- Assign the narrowly defined warning tasks to
an organization of 1-2 dozen people, depending
on the definitions used. In effect this is to
keep or reinvent the (SWS). In general, this
gives a greater Community flavor to the exercise
without interrupting the good work that Defense
is doing, but it is more costly and runs the
risk of becoming moribund. If one goes this
route the specific arguments for and against
keeping the SWS as an entity are:
-- For -- The SWS is the most significant
Community structure linking various
agencies, notably CIA and Defense, in
the warning field. It has value as
a symbol of the DCI's continuing com-
mitment to helping Defense solve what
it sees as its warning problem.
-- Against -- The SWS is widely perceived
by analysts as moribund and irrelevant.
It will carry this image into any new
-20-
Approved For Release 2004/08/16 CREIA; RDP81 B00493R000100050001-7
 SECRET
Approved Fo $elease 2004/08/16 : CIA-RDP81B00493,R000100050001-7
warning program, in effect, requiring it
to live down its reputation, something
a new entity would not have to do.
Furthermore, dispersing the SWS would
ease the problem of providing the po-
sitions for a community network of
Agency warning referents.
--- Assign all warning tasks to Community line
organizations, i.e., use the same procedures
for all warning problems:
Pro -- Properly challenged and disciplined
line organizations can do the whole
job, and at much less expense than
other approaches. Left to its own
Defense will provide structured ap-
proaches to the most serious warning
problems.
Con -- This approach is most likely to fail
because of ingrained habits of thought.
27. Two of the essential functions of a warning system
are the discipline by which the Community is kept sensitive
to the warning mission and the challenge to complacent analy-
sis that comes from "thinking ominously." Note that these
are needed for broad warning issues as well as narrow. The
question is what organizational instruments to use for these
functions in what combination. (Models A-E present a va-
riety) The choices are:
-- Line managers. This is the least expensive
and is integral to the chain of command.
Whatever additional arrangements are made,
managers will always be responsible. But
relying on them alone provides no indepen-
dent check. Moreover, managers have many
other concerns and can be as much prisoners
of conventional thinking as their analysts.
-- NIOs. This is also inexpensive and not
disruptive of normal operations. NIOs are
in a better position than managers to see
SECRET
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
 SECRET
Approved P Release 2004/08/16 : CIA-RDP81BO0+4 BR000100050001-7
the forest rather than the trees. They
can provide a low-key way of keeping the
system on its toes. But they are also
very busy and can become wedded to lines
of analysis. Moreover, the informality
of the NICE approach risks uneven applica-
tion.
-- An "ombudsman for warning." The responsible
staff officer under the DCI can serve this
function. He will not cost much, but he can,
again in an informal way, bring his weight
to bear at key points in the system. He
will be spread very thin if substantive re-
sponsibilities are added to his managerial
ones, however, and he will have to be an
individual of extremely broad ability and
experience. A court of appeal standing
outside the chain of command is likely
to create frictions.
-- An operating staff. This would extend the
ombudsman concept by providing the resources
for independent analyses. It would provide
a stronger discipline and challenge, but it
would be quite expensive and would cause great
friction by appearing to compete with line
current intelligence. Many of the arguments
relating to the SWS are applicable here.
B. Crisis Management Issues
28. As noted above, the warning issues are the pri-
mary ones at this time. But shaping the management structure
for warning and for crisis will be difficult without deci-
sions on two issues: whether we should attempt to resurrect
the "national task force" and where you locate your task
force in crisis. Both should be settled before any major
crisis hits us.
29. We assume on the basis of experience that you
will need a task force of analysts to prepare briefings,
periodic situation reports, and assessments for you and
your primary customers. (Under the new organization such
a task force will, of course, contain collection tasking
elements as well.) You can do this in four ways:
-22-
Approved For Release 2004/08/1?g.Cq-RDP81B00493R000100050001-7
 Approved FRelease 2004/08/16: CIA-RDP81B004W3R000100050001-7
A truly "national" task force, jointly manned
by the Community agencies and producing a
single periodic situation report for the en-
tire national security establishment.
Pro Provides the single comprehensive situa-
tion report that NSC has in the past
requested.
Maximum concentration of expertise.
Fully integrated Community effort.
Con -- Departmental imperatives mean each
agency will produce its own sitrep
anyway.
-- Location a problem: some analysts will
be deprived of files and support.
No agency likely to give up analysts
when it needs them the most.
Command and coordination difficult,
threatens loss of timeliness.
-- An NFAC task force with INR & DIA representa-
tion with some interagency participation.
Pro -- Provides NSC with national sitrep in
substance if not in title.
NFAC has balanced political, military,
economic capabilities across the board.
-- Fully under DCI control, but able to tap
other agencies.
-- No disruption of existing organizations
and support arrangements.
-- Timeliness not a problem.
-- Independent of policy considerations.
Approved For Release 2004/08/16 -CIA-RDP81 B00493R000100050001-7
 SECRET
Approved Fit' Release 2004/08/16: CIA-RDP81 B0043R000100050001-7
Con -- Departmental sitreps will still be issued.
-- Coordination may be difficult.
-- Liaison arrangements heavily dependent
on personalities and communications.
A variant on the above, centered on INR or DIA
rather than NFAC. The arguments are the same
except for:
Pro More "Community" in nature.
More closely in tune with diplomatic
or military operations (but not both).
-- If in DIA, stronger military input.
Con -- Did not work well when tested in DIA.
Two masters problem--not fully respon-
sive to DCI.
-- DIA cannot supply political or economic
input, State cannot supply military and
weak on economic. Both thin on low
priority regions.
An NFAC task force issuing an independent NFAC
situation report. Arguments same as for second
option except:
Pro -- Least messy administratively.
Most timely and responsive.
-- Tested and effective.
-- NFAC is sole organization dedicated to
support of NSC.
Con -- Preeminence of CIA's departmental sitrep
will depend entirely on its being superior
to others.
-24-
Approved For Release 2004/08/1~9C9 ; RDP81 B00493R000100050001-7
 SECRET
Approved I t Release 2004/08/16: CIA-RDP81 B06"3R000100050001-7
-- No pretense to national status.
-- No access to other agency contributions.
Lack of coordination may lead to errors.
30. The task force report went at some length into
the location question and summarized its analysis in the
attached table..
25X1
Approved For Release 2004/08/16 : CIAi00493R000100050001-7
 25X1 Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7
Next 2 Page(s) In Document Exempt
Approved For Release 2004/08/16 : CIA-RDP81B00493R000100050001-7