THE FUTURE OF NATIONAL ESTIMATES

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP93T01132R000100030022-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 27
• Document Creation Date: December 23, 2016
• Document Release Date: May 9, 2012
• Sequence Number: 22
• Publication Date: May 18, 1984
• Content Type: MEMO

Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
The Director of Central Intelligence
Washington, D.C. 20505
National Intelligence Council
MEMORANDUM FOR: Director of Central Intelligence
Deputy Director of Central Intelligence
Chairman, National Intelligence Council
FROM: Hal Ford
National Intelligence Officer At Large
SUBJECT: The Future of National Estimates
NIC 02989-84
18 May 1984
1. As a practictioner, observer, and critic of the national estimates
business since 1951, in and out of CIA, I believe strongly that certain
fairly substantial additional changes have become necessary in this business
if national estimating is to make the impact it deserves in tomorrow's
world. This memo examines problems which will increasingly beset the
estimate-policymaker relationship, and offers certain recommendations to
meet that more troubled future environment.
2. My chief observations/recommendations, as spelled out in the body of
this memo, are in brie :
That in some respects the coordinated national estimate has become
an outdated artform in the heavy competition for consumers'
attention -- in a world and a policymaking milieu increasingly
affected by pressures of complexity, time, and disorder.
That certain types of coordinated national estimates remain highly
necessary and should be produced, but that the NIOs, the A/NIOs,
and the NIC's Analytic Group (AG) can better serve the interests of
policymakers by continuing to increase that proportion of national
estimating which takes the form of less formal memos, think-pieces,
face-to-face encounters, new methods of communicating estimative
judgments, and so on.
That the ke to the quality of written estimates is -- and will
continue -- the quality of the drafters; that the practice of
Dave _
CL BY SIGNER
DECL OADR
t
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
borrowing drafters on an ad hoc basis from other offices has proved
a mixed blessing; that the best system yet devised for producing
the bulk of estimates is a cadre of elite, experienced estimates
officers concentrated in the estimates office staff (NIC, at
present); and that to these ends something like the present AG
should be substantially upgraded in size, stature, and recruitment
base.
With no disrespect to Bob Gates' heroic dual performance, that the
production and impact of estimates can be best maximized where the
chief estimates officer (C/NIC, or however titled) holds that
position as a full-time job, and is himself/herself a figure of
national reputation who is a hard-headed thinker/doer.
That many additional changes -- spelled out below -- are also
needed to improve the utility of future national estimating. These
encompass matters of purpose, format, procedure, media, and
marketing.
3. An increasingly difficult future market for national estimates:
-- The always difficult market for estimates is going to get worse.
The producers of estimates, up and down the chain of command, must
recognize more clearly that their efforts will face heavy
competition indeed for the time and attention of senior
policymaking consumers. These key targets of ours are the very
officers who have the least time and energy to absorb our wisdom.
They carry their own NIEs around in their heads. They often feel
that they do not need us, especially in fields where general
knowledge is plentiful, but unique augmenting intelligence is thin.
-- There is going to be no automatic market of expectant consumers,
just waiting for our estimative insights before they proceed to
policy decision. Dispassionate estimates are going to be up
against advocacy, with the latter having the advantage of always
being simpler and more seductive. And in particular, our estimates
will not encounter a ready market on those occasions where their
portraits of the world are not congenial with policymakers' own
images or commitments.
-- The expanding hazards to estimates' impact will be both foreign and
home-grown. Tomorrow's world will bring not only the growing
weight of the Soviet global challenge, but increasingly more
volatile threats to US interests from instabilities in the Third
World and elsewhere. Such rising disorder will create a more
difficu 1 ~olicymaking milieu. The demands of meeting pressing
crise as lways produced what past valuations have correctly
termed a stranglehold" by current intelligence, to the deteriment
2
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
of sufficient consumer or producer attention to longer-term -- and
often more serious -- problems. This situation will intensify as
policymakers are beset by a rise in the number, complexity,
insistence, and time-squeeze of world problems.
Accompanying this trend will be certain new hazards to estimates
arising from improved White House and other operations centers such
as that of Richard Beal's. These efforts will be good/bad: they
will tie intelligence to policy on a more immediate basis, but at
the same time may damage decision making by surrounding senior
policy officers with facts and judgments which in some instances
are more high-impact than accurate or meaningful.
For many reasons, hence, there will be more disorder in tomorrow's
world and tomorrow's policymaking -- and, consequently, a greater
gap between the very rational purposed theory of national
estimating on the one hand, and the more haphazard practice of
policymaking on the other. This means that tomorrow's national
estimating will have to be damn good in quality and utility, on and
beyond recent improvements, if it is to justify the time, talent,
energy, and taxpayers' money spent on its preparation.
4. The case for fewer interagency national estimates and more national
estimating:
The case still exists -- more than three decades since the creation
of the NIE art form -- for the traditional purposes of certain
national estimates. Those purposes, as expressed by then-DM
Bedell Smith, sought in the national estimates an authoritative
interpretation and appraisal that would serve as a firm guide to
policymakers and planners, a disinterestedness above question, the
collective judgment of the highest officials in the various
intelligence agencies -- hence commanding respect throughout the
government as the best available and most authoritative body of
estimative judgments. These considerations still apply for many of
the basic studies, such as the NIE 11-3/8 series, where an NIE
serves as an agreed reference point for key planning; and for
evaluations of certain other crisis or troubling situations of
pressing importance to the United States where authoritative,
dispassionate basic assessments may be in short supply.
But in the case of many of other types of national estimates, the
institutions of orderly policymaking for which estimates were
designed originally to serve have long since disappeared. Apart
largely from long-range military planning, policymaking takes place
much more on the run. The best step the estimates business has
taken to meet this changed circumstance is the creation and
*IAC-M-1, 30 October 1950.
3
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
strengthening of the NIO system. Well and good: through various
means the NIOs have moved out smartly into this policymaking
scene. But the ties of estimates and policymaking are still
somewhat hit-and-miss, with no systematic match-up, and with the
time and talent of senior NIC officers overly drained off in often
feckless coordination.
There continues to be a sizable gap between the theory and the
practice of the coordination process. At the representatives'
level there is often a lack of individual candlepower, geniune
expertise, and actual authority to represent the Principal. With
some exceptions, representatives tend to defend prior established
positions, or just insure that nothing too objectionable gets in
the text, or just pass the buck along to the Principals. There is
strong reluctance at many representatives' meetings to take clear
dissents, or to undertake new kinds of inquiry or lines of march,
or to venture out beyond demonstrated intelligence at hand, or to
judge the possible consequences of possible future developments.
These drawbacks are reduced, the better and stronger the texts, and
the stronger and better the NIO Chairman. Often the coordination
process improves an estimate's precision and introduces new
subtleties into the text. Drawbacks nonetheless persist, and so
create many other situations where the final coordinated draft that
emerges is essentially that which entered the reps' arena, only
less sharp, less clear, of less utility -- and much delayed.
There have been worthwhile efforts to increase the participation of
Principals in the estimative process. Again, well and good, and
the more such continuing pressure on them the better. But,
realistically speaking, the fact that most Principals are
essentially managers is always going to make the outcome of NFIB
meetings largely the result of given DCI's and whatever assorted
creative personalities happen to attend the particular session,
rather than the collective wisdom foreseen by General Beedle Smith
and his original IAC.
Given all these limitations on national estimates, there is a
strong case to be made that the NIC (and future central estimative
offices, whatever their title) can best serve policymakers by
conceiving of themselves more as national estimators rather than as
just the producers of national estimates. This means that the
NIC and the can an should be manned by the most sophisticated,
broadly experienced officers that can be gathered together; and (2)
that these NIC officers not dilute their contribution to national
estimating by having to spend too great a proportion of their time
grinding out coordinated NIE packages.
4
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
Constructive critijs have long warned estimators of the dangers of
over-coordination. What have been often the most valuable
inputs made by senior estimative officers over the years have been
sharp ad hoc or in-house studies which break new ground, point out
new developing world threats or opportunities, question
conventional wisdom, examine the consequences of contingent
developments, or otherwise give policymakers more direct, focussed
assistance than can the necessarily more ponderous estimates --
even the recently improved fast-track variety. NIOs, A/NIOs, and
AG members are in the best possible spot to contribute such
insights, and should be encouraged to continue to enlarge the
proportion of such efforts, checking carefully in each instance
with DDI or other appropriate specialists, and indicating clearly
to the readers the status of the views being presented.
Policymakers would be well served also if, on occasion, memos of
comment were offered on such think pieces by individual NFIB
Principals or other senior intelligence and policymaking officers.
NIOs, A/NIOs, and AG officers, if freed somewhat from the sizable
paper-shuffling demands of coordinating and producing formal
estimates, would have more time also to assist other senior
intelligence officers in guiding collection and in devising new
means of communicating estimative findings, in addition to that of
the printed page. Impact on the faster-moving policymaking world
will require much more in the way of video, graphics, face-to-face,
and other measures. Also more emphasis, see below, on marketing
and follow-up.
In all such cases of estimating by means additional to national
estimates, the payoff must of course remain on the quality and
utility of estimative assistance to policymakers, not on the
quantity of NIEs or other estimative pieces being produced.
5. The key importance of an estimate's drafter:
-- Another clear fact which three decades of US estimates experience
has demonstrated is the absolutely primary importance of the
particular drafter to that finished estimate's quality and
usefulness. Where initial concept and drafts are only so-so, or
worse, they not only clog up the estimates schedule but often
*For example, this ancient but still apt recommendation, from a senior
CIA officer, 1957: "The sum and substance of what I have been saying is
that the US national security system would be better served if the
Intelligence Community took a less vigorous view of the meaning of
coordination and substituted more informal techniques of consultation."
5
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
remain relatively impervious to subsequent tinkering and
re-drafting.
Where drafters are top-rate there is no problem. But the record is
not one of unblemished success, now or in the past. Traditionally
the toughest cases exist where the drafter proves mediocre or
poor. It is not always easy to know in advance whether an untested
drafter will do a good job of preparing an estimate: some good
current intelligence officers, for example, have put facts and
chronology together in an "estimate," but one which to the consumer
has no so-what. The writing of estimates calls for distinctive
experience and breadth, as well as distinctive skills in
conceptualizing, organizing, and presenting an estimate's findings.
The 1974-1980 experiment which required NIOs to scrounge estimates
drafters as best they could proved a failure -- one recognized in
the decision to reorganize the NIOs into a NIC, supported by its
own AG. Since that time the drafting situation has improved
somewhat, but because of the AG's small size and the many demands
on the time of the NIOs and A/NIOs, the majority of'estimates still
has to be farmed out to other offices.*
This farming out of drafting assignments involves various
problems. Outside drafters do not belong to the NIOs. They are
not answerable to NIC discipline or standards. They are sometimes
physically separated from the NIO chairman, even across town. NIOs
don't always get the drafting stars they seek, but have to settle
for those the parent offices make available. In some host offices
the drafting of national estimates is not treated as part of a
career-enhancing pattern, but an external chore. Drafters are
caught between the demands and views of their own offices and those
of the NIO. In result, enthusiasm, priority, quality, and an
estimate's usefulness all suffer.
Some farming out of estimates must of course continue. This
certainly applies for many of the complex military estimates where
outside-the-NIC analytic offices have produced many good drafting
teams. The same applies for those particular occasions where the
dimensions of a given estimative chore happen to fit the analytic
culture well, and where the host offices do ante up first-team
drafting talent. But there are limits to such practice, including
distinct limits on how much burden NIC projects should exert
especially on DDI production offices' own responsibilities.
The answer: an increasing proportion of coordinated estimates and
in-house pieces can best be done by an experienced AG of
*The 1983 record: 32 interagency estimates were drafted by DDI
officers; 24 by NIC; 9, DIA; 5, INR; and 8, joint.
6
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
strengthened proportions, the best type of system yet devised for
developing creative estimates drafters. A group encompassing such
breadth, intellectual leadership, and skills can also constitute a
high-class drafting pool for special ad hoc DCI and C/NIC chores.
This cannot be done well, however, by the present AG. As initially
organized by D/NFAC in early 1980,* this group was to consist "of
about 20 officers;" those officers were to draft "the bulk" of
coordinated estimates; they were in addition to "initiate ad hoc
estimative memoranda for NIC discussion and futher disposition;"
and rotational tours in the AG were to be an "important element in
the career planning of NFAC offices." None of these situations
exists at the present time. The AG now has only 11 professional
slots. Its members draft only coordinated estimates, not think
papers as well. CIA chiefs do not willingly provide the AG their
best officers for rotation tours but understandably husband them
for their own offices' purposes. Nor, except for military hardware
questions, is there much sophisticated drafting talent available in
the Intelligence Community -- we have had one such tour in the AG
which was successful (NSA), one which proved mis-cast, and one
(DIA) up-coming. The record has also been mixed in drawing top
talent into the AG from academia, etc., where this path also
entails special bureaucratic hazards.
In short, if intelligence is to offer the maximum possible support
to policymaking, it must have an estimates cadre of the best brains
and effectiveness in town. This did obtain at certain times in the
past, witness the wealth of talent represented by such former
estimates staffers as Hyland, Billington, Komer, Maury, Cline, W.
Bundy, Carver, Byrnes, R. J. Smith, A. Smith, D. Graham, Huizenga,
Clarke, Whitman, Chet Cooper, and many others. The principal
reason such talent had been made available was that the estimates
office was initially conceived to be the heart of the CIA and of
the national intelligence machinery,"* and early DCI's made sure
that the estimates office got assigned the elite drafters it
required. I submit that something like this concept of an
estimates drafting group is required, or at least something
approaching the AG as initially envisaged in early 1980, if the
estimates business is not to continue bumping along, doing a fairly
good job, but not living up to the potential it could contribute.
6. The need for a full-time C/NIC:
-- The C/NIC is a more than full-time job in itself. The Chairman
must furnish intellectual leadership, get the most out of his/her
*NFAC Notice No. 1-19, "Responsibilities and Structure of the NIC
Analytic Group," of 30 January 1980.
*IAC-M-1, 20 October 1950.
7
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
officers, administer the office, and relate actively to senior
members of the intelligence and policymaking communities. This
latter requirement is of paramount importance inasmuch as
estimates, being somewhat free-will offerings, will always have
greater impact the more the estimators are known commodities to the
policymakers, not faceless officers somewhere across town. To
important degree the regard in which given estimates are held rests
on the personal respect in which their producers are held. This
applies of course to all the members of the NIC, but in particular
to C/NIC. He/she must have the opportunity to spend needed time
with senior officers around town (and with the country's best
brains, wherever) before, during, and following the preparation of
estimative support -- and so multiply the impact of the estimates
effort.
-- Although there have been excellent chiefs of the estimates office
who came there from CIA careers, there will generally be an edge in
stature, contacts, and impact -- all other things being equal --
where C/NIC is a scholar or official of national reputation. In
short, future NIC's can be most effective when they have something
like latter-day Bill Langer's in charge.
7. Additional recommendations for improving the quality and impact of
estimative products. Here purposely avoid familiar criticisms many o ers
have made, and confine my points to capsule presentations. In brief, there
is need for the DCI to direct that much greater attention be devoted to:
The marketing of estimates -- by the DCI, C/NIC, and NIOs alike.
The most rewarding measures involve personalized intervention at
various stages of key exercises, before and after their
production. There is some of this now, from time to time, but
unless pressed much more, our finished products will continue to
tend just to pile up, undifferentiated from other mail, on the
desks of special assistants and other filters. There needs to be
much greater consciousness that our work is not completed at NFIB.
Otherwise we short-circuit the process and the purpose of
estimating.
More re ularized evaluation of estimates. To date this has been
confined to sporadic ad hoc efforts, aimed generally at examining
"failures." Fuller and more regular evaluations, conducted by
senior, objective groups, could transmit back much-needed guidance
as to what has and has not been accurate, useful, etc. This cannot
be done by just reading stacks of old papers, but must involve
considerable interviewing, the building of personal contacts with
consumers, and demonstrated evidence to them of the worth of such
inquiry. Some estimates could benefit by making a review of
previous judgments on the same topic an explicit part of their
content.
8
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
More attention to collection re Third World developments. Here is
where most of the action is, an were the prime detonators to
world peace are. The Intelligence Community (especially State)
must be prodded from on high to get US missions out of their
cocktail cocoons and into their host societies, so that blindsided
analyses and estimates do not inflict more self-harm on US
policymaking.
More attention in estimates to factorin out the respective
indigenous - external Coninunist ingredients in Third or of
sots. Such crises are of course of enormously greater danger to
interests where Soviet or other hostile elements are at work in
the picture. But US policymakers have paid dearly in the past for
their relative ignorance of those basic forces in certain world
settings which create the local pro-Communist and without whose
remedy many US well-intentioned policies will go unavailing.
Less emphasis on redictin events, more on depicting forces and
trends at work in given estimative situations.
More estimative emphasis on giving policymakers handles: that is,
pointing up opportunities as well as threats, and erentiating
between those forces in a given picture which seem inexorable, and
those others that may to x degree be amenable to US or other
friendly remedy.
Being less shy, in estimates, in suggesting opportunity handles to
o is ma ers. of trying to make Policy, but not stopping either
wit just telling the consumer that he/she faces a hell of a
situation in Ruritania.
More contact b estimators with the country's best brains outside
of professional intelligence ranks. Contact with outside experti
and consultants remains sporadic. More is needed, and on a fuller,
more systematic basis, to avoid certain stultifying effects
Washington localitis can involve.
Much more effort b and on behalf of the estimators to know the US
Blue element much better -- and making sure that such knowledge o
t e ingredient is ground into analyses and estimates of foreign
situations.
Better appreciation among analysts and estimators that the, too,
not only the policyma ers, must keep alert to the distorting
influences of prior belief.
9
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 i
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
CONFIDENTIAL
Finall , a licable in relation to all the above, a fierce
determination estimators to tell it like is: tfi-af--Ts, the
necessity to give our consumers the fullest and most objective
analysis/judgment possible -- without regard to the policymakers'
particular preconceptions, commitments, or sensibilities. It is
the job of estimators to tell the truth, not to make our customers
happy. Otherwise we will just be spending taxpayers' money to help
policymakers deceive themselves, on occasion, about how well things
are going in Vietnam, or Iran, or Lebanon, or wherever.
8. I will be pleased to learn your reactions to this memo's
observations/recommendations, and to discuss these matters further.
/ 6-94c/-,
10
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
SUBJECT: The Future of National Estimates
DCI/NIC/NIO/AL/HFord:ps.
(18 May 84)
STAT
Distribution:
Orig - DCI
1 - D DC I
1 - EXDIR
1 - SA/DCI
1 - ER
1 - C/NIC
1 - ADDI
1 - VC/NIC
1 - SRP
1 - Each NIO
1 - AG
1 - Ford Chrono
11
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 i
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
TRANSMITTAL SLIP
DATE
TO:
NI0/EA SUBJ
ROOM NO.
BUILHDIINsG
REMARKS:
FROM:
ROOM NO.
I
BUILDING
EXTENSION
IXw 241
WHICH MAY BE U30.
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
yam',
~ f.
The Director of Central Intelligence
WuWngton, D.C. 20505
National Intelligence Council NIC #00362-84
18 January 1984
MEMORANDUM FOR: Robert G. Gates
Chairman, National Intelligence Council
THROUGH . Charles E. Waterman
Vice Chairman, National Intelligence Council
Herbert E. Meyer
Vice Chairman, National Intelligence Council
INFORMATION : All National Intelligence Officers
National Intelligence Officer for East Asia
SUBJECT : Simplifying NIC Procedures
For the NIO, the process of managing an Estimate has become nearly as
time consuming and taxing as the much more important task of paying attention
to the substance. The process of preparing an Estimate includes:
-- writing countless memos to transmit various drafts,
-- circulating drafts to representatives in a hurry.
Too Many Memos
Let us first consider the required memos along the Estimates trail. The
Comment column suggests improvements in the existing system.
Memo Comment
A memo is sent to C/NIC and the DCI The NIC buck slip with a blue stripe
asking for approval of the topic. (hereinafter buckslip) should be
sufficient to convey three or four
sentences of information explaining what
the topic is and who requested it.
After the TORs are received, a memo This same information could be conveyed on
is prepared for VC/NIC to send to a buckslip.
SRP asking for their comment.
CL BY SIGNER
DECL OADR
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
SECRET
Simultaneously, a memo is prepared
for the DCI asking approval of the
TORS.
Next, a memo goes to NFIB principals
asking for the appointment of
representatives.
After the NIO has reviewed the
draft, a memo is prepared to C/NIC
and the SRP asking for their
comment.
The DCI receives a memo conveying
the draft and asking him to approve
its submission to reps for
coordination.
With DCI approval in hand, a memo is
then sent to reps calling a
coordination meeting.
After coordination, a memo conveys
the coordinated draft to the DCI for
his final approval.
Prior to convening of NFIB, a memo
is prepared to the DCI providing him
with talking points for the NFIB
meeting.
Ditto. A buckslip should suffice if it
really is thought necessar to obtain DCI
approval o Rs. Persona question
th s step.
Instead of sending this memo by individual
envelope, we could transmit the text by
Wang to the IC staff for circulation by
Wang to principals.
A buckslip should be sufficient.
Is a memo necessary? A buckslip should
suffice.
Unnecessary. A memo has already gone out
to NFIB principals asking for the appoint-
ment of reps. At this juncture, all the
reps need to know is when to come to the
meeting. This should be accomplished by
Wang to 0 for onward "Wanging" to
principals asking them to inform their
reps.
A buckslip should do.
In practice, the NIO who presents the
Estimate does the talking. The DCI has
little use for such a memo. A buckslip
should do, or better still, this step
should be a iminated a together.
This proposal eliminates most memo writing. Any secretary will confirm
that there is a vast difference between typing a blue stripe buckslip and
preparing a letter-perfect memorandum for the DCI. In addition, this proposal
eliminates the problem of addressing up to a dozen separate envelopes for two
of the necessary memos -- one to NFIB principals asking for the appointment of
representatives and the other to representatives givin the time and date of
the coordination meeting. By "Wanging" both of these for onward
electrical transmission, the addressing of individual envelopes twice is
eliminated. will not retransmit memos; they would have to be
instructed.
2
SECRET
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
Moving Drafts to Reps
Second, let us consider the problem of moving drafts to representatives
prior to coordination or between coordination meetings if there are to be
several. Ideally, the interagency courier system is available to move these
drafts. But experience has shown -- and most NIOs will confirm -- that this
system does not work-well enough to suit our purposes. Often we are in hurry,
either because we are moving a fast-track or at least a quick paper, or
because the courier system simply breaks down and papers sit in mailrooms.
_
for
The most sensible remedy is to move the drafts by the Wang system F
onward "Wanging" to principals to ether with a buckslip asking principals to
pass the draft to their reps. will not handle drafts; they
would have to be instructed. Nor are principals prepared for this change;
their role would have to be spelled out and their cooperation enlisted.
To sum up, the problems of the process now match the problem of
substance. Something must be done. At a -minimum, memo writing should be
greatly curtailed; at a maximum, our old-fashioned, inefficient way of moving
drafts around town should be replaced by expanded use of the Wang system.
25X1
25X1
3
SECRET
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 i I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
CONFIDENTIAL '"
May 1983
The National Intelligence Council and the Estimative Process
In the US Intelligence Community's division of labor, the National
Intelligence Council is the unit charged with overall responsibility for
managing the production of interagency estimates, but each component of the
Community contributes to and plays an active part in the process.
The following paragraphs describe how estimates are produced now, with
particular emphasis on the changes that have been introduced in the process in
recent months.
Production of an estimate begins with obtaining the DCI's approval for
its preparation. (An estimate may start from the request of a senior policy
officer or the DCI himself, from a standing requirement for periodic
assessments, or from an internal NIC proposal. In practice, relatively few
estimates stem from NIC proposals; 16 of the 59 formal estimates produced in
Calendar Year 1982 were NIC-initiated). A new estimative project is assigned
to one or more of the 14 National Intelligence Officers (NIOs)* to manage. As
Chairman of the paper, the NIO finds a suitable drafter, engages the rest of
the Community in the production effort, sets the timetable, and presides over
the paper's completion, review, and coordination. The first significant step
in this process is to develop Terms of Reference (TORs) and a Concept Paper
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
for the project, seeking assistance for this purpose from a variety of
sources, including, where appropriate, the requestor himself. In 1982, in
order to engage the senior leaders of the Intelligence Community more fully in
this crucial step of defining the scope and direction of each estimate, we
began sending draft TORs and Concept Papers directly to the principal members
of the National Foreign Intelligence Board (NFIB). Heretofore, TORs had been
drafted, reviewed, and coordinated at the working level, and the personal
involvement of the Community's senior officers early enough to have an impact
on the shape of an estimate was a rare occurrence. We believe that starting
the review of an estimative outline at senior levels and progressing from
there to coordination at the working level is a better, potentially more
fruitful way to begin, and we now do this routinely.
Once the bare bones of the project are agreed upon, the drafting stage
ensues. In choosing a drafter, the NIO has the entire Intelligence Community
to draw on, and he normally seeks the best analyst for the job at hand that he
can find, regardless of agency. Some of the larger papers, particularly the
complex military ones, are written by interagency groups, and others are
produced on the basis of contributions from more than one agency, but mdst are
written by a single drafter, working under the guidance of the NIO. In 1982,
34 of 59 estimates were written by CIA Directorate of Intelligence drafters,
14 by NIC analysts, five by DIA, four by INR and two by more than one agency.
When a first draft of an estimative text is achieved that is acceptable
to the paper's Chairman and to Chairman, NIC, formal and informal comments and
advice on it are solicited from a variety of sources, including the DCI's
2
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
l.vrlr lucn 11/1L
Senior Review Panel (a small, independent staff of distinguished senior
officers with long experience in government), other NIOs, expert analysts
inside the Intelligence Community, outside consultants, and, time permitting,
appropriate US experts in the field. Once their suggestions have been
factored in, a revised draft is prepared and sent to the DCI for his review
before the coordination process begins.
Coordination usually involves meetings of the representatives of all
agencies of the Intelligence Community, although on occasion, agencies with no
expertise on a particular subject will opt out of the discussions. The real
purpose of the coordination process is to enrich the paper by examining the
facts from several different organizational and judgmental perspectives.
Although there is an inevitable and inherent tendency in any coordination
process to water down or negotiate away differing points of view, we try very
hard not to bury or smooth over genuine points of disagreement in
interpretation of the evidence or key conclusions. What we strive for is a
process in which minor or inconsequential differences are talked out and
resolved and major ones are illuminated and more sharply defined. We succeed
in this only imperfectly and are very much aware of the need to work
constantly to avoid producing "least-common-denominator" estimates.
The draft text, once the coordination sessions at the working level are
completed, is then resubmitted to the DCI for his approval to schedule it for
consideration at a meeting of the National Foreign Intelligence Board
(NFIB). The DCI issues estimates with the advice of the Board. Each NFIB
Principal has the right to dissent from any of the findings of the paper.
3
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
CONFIDENTIAL
Again, we have been trying in recent months to invigorate the process by
urging the Principals to become more deeply involved in the development of
each estimate from the very beginning. In the past, the Board rarely got
involved with an estimate until the very last, when it generally ratified a
paper whose language (including dissents) had already been worked out at a
lower level. Even with greater involvement in the estimates by Board members
these days, their routine concurrence in the final draft is by no means
assured, even when their representatives have already signed off on it. Board
action remanding a paper for more work, or a Board member's tabling a dissent
or a significant change now takes place at the NFIB meeting more often than it
did formerly. We regard this as healthy and we think it is likely to enhance
the final product and make it possible to give the policy reader a truer
picture of the Community's views on important issues.
Any process involving as many actors and as many steps as those outlined
here must be seen as fairly ponderous in terms of the time consumed from start
to finish. To combat that problem, we have introduced, and continue to
refine, so-called "fast-track" procedures aimed at shortening the interval
between a paper's inception and its completion. We use these fast-track'
procedures -- which essentially rely on heavy use of the secure telephone
system and doing as many steps as possible simultaneously instead of
sequentially -- only for papers where there is a real need to receive the
Intelligence Community's views fairly quickly. Whereas we did only a handful
of fast-track papers in 1980 and 1981, we did 14 last year. Our "record" so
far is three working days from start to finish. Producing papers this quickly
4
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 i
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
V CONFIDENTIAL
requires a degree of teamwork and puts a strain on the machinery that we
cannot invoke too often, but it permits us to weigh in with our combined best
judgments on a fast-moving situation quickly and effectively.
Most SNIEs are now produced in periods of up to a month or two, but
National Intelligence Estimates (NIEs) and Interagency Intelligence Memoranda
(IIMs), our more deliberately-produced estimates, usually require several
months to complete -- or more, in the case of lengthy and complex military
assessments.
Thanks at least in part to our fast-track production procedures, our
capability to respond to requests for assessments in support of specific
decision points in the policymaking process has been steadily growing. The
vehicle we most often use in instances of this kind is the Special National
Intelligence Estimate. Last year, 31 of our total of 59 formal estimates were
SNIEs, as opposed to 18 in 1981 and 7 in 1980. While not all SNIEs are
targeted for specific policymaking needs, many are.
5
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
1. NIEs, SNIEs and IIMs are all members of the
category of estimates. One estimates when one does
not know; and all estimates are analytical constructs
arrived at on the basis of usually inadequate informa-
tion and of assumptions that govern the interpretation
of this information. Since estimates are essentially
inferential, and usually speculative regarding the
future, and since no one inference or speculation is
necessarily compelling, it is sometimes neither possible
nor desirable to produce a unanimously agreed estimate.
While effort should be made toward consensus, a set of
alternative estimates should be presented and argued
whenever irreducible uncertainties preclude agreement.
Since estimates are hig1}ly sensitive to analytical
assumptions about the real world with the help of which
information is interpreted, it is vital that key assump-
tions be spelled out and, if necessary, defended in the
presence of conflicting information. It is especially
important that this be done in the event of estimative
disagreement.
2. Given these common properties, the three art
forms may be defined as follows:
a. An NIE is an estimate of a foreign
situation that impinges importantly on US
interests and is relevant to the formulation
of important choices for foreign and national
security policy, and which, in these terms,
conjectures about future developments of this
situation. It is normally incumbent on an
estimate to illuminate how US interests are
affected by the situation and how, in turn,
this situation and its development may be
affected by US policy. If estimative
uncertainty cannot be resolved, it is also
incumbent that the paper set forth alternative
estimative conclusions. NIEs are in principle
products of the entire Intelligence Community
and are issued by the DCI with the advice of
NFIB.
DEnidATI' E C. L EY SIGNERS
c EEC! ME; W Cry 2 Oct 1985
DERIVED FROM
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
 .I I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
FZ
CONFIDENTIAL
b. A SNIE differs from an NIE in that it
is relevant to a specific and urgent policy
problem, and that it is therefore shorter,
prepared more quickly and coordinated with the
Intelligence Community with dispatch. It is
issued by the DCI with the advice of NFIB.
c. An IIM differs from an NIE in that,
although requiring the endorsement of the APPROVED StI
Intelligence Community, it is issued without Gu te.. Ok
NFIB concurrence, deals with policy issues of pkc-swTeo n-r
lesser significance, and may be either an NF16
assessment of some current situation or an
intelligence estimate.
As indicated, the choice and application of all
three art forms cannot in some important respects be
determined by precise rules. This holds true especially
of the choice between IIMs and NIEs (and SNIEs), and on
the extent to which an-estimate will engage questions of
US policy. But while these choices can be settled only
ad hoc, they are too important to be made casually or
without review. Ultimately these are, of course,
management decisions.
d. -l A -- SHORT- EST1MkrE5 PRO QUCEO Qvle. Ut WHEQ A
MORE "AVVL pAPE2 LS OMAPPOPgOT& OUE
TD XQS(TWE WJP-CES U M ITCI) WISSCWuA noun
AIM~TEO &)FI PAQTItOAi Oz
APPROVED 4 CENIC -- Mhq PJE REV!EuOEO 13~
TuEoat
N!elul -- iQF:o m,*t N( A*MOVEA/AA WLT4 LlMtr&
Ot? IJO C- TUN4L IAA rIGAr4noN1
A(tROUCO
ey CIIJIC
CONFIDENTIAL
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
SECRET
The Director of Central Intelligence
Washington, D.C. 20505
National Intelligence Council NIC #03324-R4
7 June 1984
NOTE FOR: Bob Gates
FROM: Dave Gries
SUBJECT: NIC Collegiality
You asked for suggestions on how the NIC could better realize its
potential as a collegial body.
First, the written product. Broadly speaking, the NIC currently produces
regional estimates and cross-cutting estimates. The former in many cases will
not benefit from collegial treatment; the latter, the cross-cutting estimates,
in most cases will benefit and you have recognized this in institutionalizing
a panel system to hring several NIOs into the process.
Under the panel system, a single NIO chairs an estimate and other NIOs
are brought in as advisors. What I am proposing is a little different. For
cross-cutting papers a collegial grouping of NIOs would have co-equal
responsibility for production of a very short paper--not an estimate--that
would probably be written by a single AG drafter and intended for the DCI.
Although a little unusual or untidy from a management point of view, the NIOs
are a cohesive enough group to work in this fashion. The prototype of course
is the original Board of National Estimates which acted collegially in
reviewing drafts submitted by a large ONE staff.
Two cross-cutting topics for short, hottom line only treatment that might
be used to test this approach are:
-- Economic reform in communist countries such as China, Hungary, East
Germany and Czechoslovakia. What does it mean and where will it
lead? Such a paper might lead the OCT to assign analytical resources
to the topic, request an interagency view or yawn.
-- Why do some developing countries develop when others don't? What are
South Korea, Hong Kong, Taiwan, Singapore, Malaysia, and to a lesser
extent, Rrazil, Mexico and Argentina doing right and what are all the
other developing countries doing wrong? Is there a formula? Do [IS
policies take account of this formula if it exists?
, Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
Second, oral presentations. The NIC acting collegially can be a sort of
think tank, meeting occasionally to address orally a preset topic. Some of
these sessions might spin off a memo from the Council to the OCI; others would
enlarge the intellectual capital of the participants. One Tuesday staff
meeting could be set aside each month for consideration of a topic of interest
to all. Discussion topics need not be cross-cut, that is, they need not draw
on the experience of each NIO, since there is good value in asking an officer
to participate in discussion of a topic in which he may have had only passing
interest. Some topics for this kind of oral discussion could include:
-- Strengthening the Intelligence Community. What could be done?
Should CIA play a more active role in recruiting analysts for DIA and
INR? Should analysts be exchanged among Intelligence Community
agencies with greater regularity? Can we convince State to increase
TNR's budget and develop a professional core of analysts?
-- Is the Intelligence Community's written product apportioned in the
same fashion as the interests of the policy community? For example,
do we overwhelm policymakers following Soviet activity and
"underwhelm" those following Third World issues? What do the
statistics show?
-- Imagery and signals intelligence play an ever increasing role in
production. Are analysts throughout the Intelligence Community
properly trained to use this information? Does CIA have a training
role in helping other Intelligence Community agencies?
Third, if not exactly collegial, the NIC can avail itself of educational
opportunities that are there for the taking. Outside speakers can be asked to
address us on a whole range of topics. Since the NIOs are inveterate
briefers, they have a lot of chits out in the government and academia and
could take charge of arranging a program of outside speakers from among these
obligated contacts. Some subjects worth exploring:
-- Kremlinology: what is the state-of-the-art? Speakers in academia
could at least provoke us on this topic.
-- Social spending and GNP growth. Among OECD countries statistics
support the notion that high social spending is the enemy of high GNP
growth. But the reverse is not necessarily true unless a range of
other conditions is met. This is a fascinating subject and has been
widely addressed in academia.
-- Third World or third rate? The six or so newly industrialized
countries notwithstanding, why have so few Third World countries
shown progress towards development? Do the political systems of the
slow growers have anything in common? Is there a political
correlation between growth and drift? Again, such general questions
are frequently addressed in academia and probably in such agencies as
AID as well.
Beyond these suggestions, it would be helpful in fosterin g a collegial
spirit if individual NIOs knew more about each other's work. One simple step
would he the creation of an Estimates library in the registry--because it is
9
SFCRFT
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
vaulted--where all NIFs, SNIFs and IIMs of the last several years could he
permanently located and arranged by subject. This would make it possible to
browse through the works of other NIOs as time became available. The current
system of circulating drafts is not productive. Which busy NIl) wants to read
a draft that may he changed at the table?
Collegiality would also be enhanced if each of us had a clearer picture
of what other NIOs were doing. Most NIOs have distinctive work styles and
unless communication between NIOs is promoted, there is no ready way to
benefit from the best points of the styles of other NIOs. This can be
accomplished by allotting 30 minutes at the large Thursday staff meeting
during which each NIO would give a two-minute report on his recent and future
activities.
cc: Herb Meyer
Hal Ford
3
SECRET
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 i
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
7
r
Chairman, National Intelligence Council
SUBJECT: Collegial Review of Estimates (U)
1. I am aware that some of you call on your NIO colleagues
for comment on estimates for which you are responsible. I am
also aware that Herb has organized periodic ad hoc groups of NIOs
to review particular estmates. (U)
2. I believe the time has come to regularize this
consultation and extend it to virtually all of the formal
products of the NIC. I would appreciate your sending me a list
of estimates and IIMs you have underway and the NIOs with whom
you are already consulting and with whom you believe it would be
appropriate to consult on each. I believe this process will be
most valuable if it begins with the conceptualization of the
estimate and proceeds through each successive phase. I urge you
to keep in mind those of your colleagues with broader
responsibilities such as Maurice, Dave Low, Ted, Fritz and Hal.
(C)
3. From now on, whenever you propose an estimate or an
estimate is assigned, you should create an NIO panel and inform
both Herb and me of the members. Your consultations can either
be individual or in meetings -- flexibility and your convenience
are important -- but I am convinced that the overall quality and
sophistication of our work can be improved by greater collegial
collaboration. (C)
4. I will expect your reports on present consultations
referenced above by COB 8 June. (U)
CONFIDENTIAL Cl By Signer
DECL OADR
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7
 I
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93T01132R000100030022-7
NOTE TO: All NIOs
FROM Chairman, National Intelligence Council
SUBJECT: Collegial Review of Estimates
1. I am aware that some of you call on your NIO colleagues
for comment on estimates for which you are responsible. I am
also aware that Herb has organized periodic ad hoc groups of NIOs
to review particular estmates.
2. I believe the time has come to regularize this
consultation and extend it to virtually all of the formal
products of the NIC. I would appreciate your sending me a list
of estimates and IIMs you have underway and the NIOs with whom
you are already consulting and with whom you believe it would be
appropriate to consult on each. I believe this process will be
most valuable if it begins with the conceptualization of the
estimate and proceeds through each successive phase. I urge you
to keep in mind those of your colleagues with broader
responsibilities such as Maurice, Dave Low, Ted, Fritz and Hal.
3. From now on, whenever you propose an estimate or an
estimate is assigned, you should create an NIO panel and inform
both Herb and me of the members. Your consultations can either
be individual or in meetings -- flexibility and your convenience
are important -- but I am convinced that the overall quality and
sophistication of our work can be improved by greater collegial
4. I will expect your reports on present consultations
referenced above by COB 8 June.
CONFIDENTIAL Cl By Signer
DECL OADR
Declassified in Part - Sanitized Copy Approved for Release 2012/05/09: CIA-RDP93TO1132R000100030022-7