INTELLIGENCE CONSIDERATIONS IN THE FORMULATION OF NATIONAL SECURITY POLICY
Document Type:
Collection:
Document Number (FOIA) /ESDN (CREST):
02437974
Release Decision:
RIPPUB
Original Classification:
U
Document Page Count:
37
Document Creation Date:
December 28, 2022
Document Release Date:
August 9, 2018
Sequence Number:
Case Number:
F-2013-02252
Publication Date:
October 21, 1959
File:
Attachment | Size |
---|---|
![]() | 1.61 MB |
Body:
Approved for Release: 2018/07/24 CO2437974
INTE,'LLIGENCE COSIDET,ATIOITS IN TI-117,
FOn'arLATIOli OF LATIOilla.
BY
rjav, Honorabl (A. riolx.ret Arlo
�
NO
'c'This is oilomca1 dc.,ct,recnt of The
Nationel Cc. frua
libstractftp cr :.!1
or 11.y pa:.1 'is ri:3T
AUTNCUZED perri!zsiN
of the Co:nrciandaci',. i'ho Nal!or,.-11 War
Presented at
The Uational War College
Washin3ton, D. C.
21 Octol,er 1959
2,
1 8 1959 RECD
Cony 01
- - -
1;7:f.:-
,
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
BIOGRAPHY
The Honorable Robert Amory, Jr., Deputy Director for
Intelligence, Centra'Intelligence Agency, was born 2 March 1915
in Boston, Massachusetts.
He received his B.A, (1936) from Harvard College and LL.B.
(1938) from Harvard Law School. He was admitted to the New York
bar in 1939; the New Hampshire and Massachusetts bars in 1946;
and practiced law in New York City, 1938-40. He served as pro-
fessor of law and accounting, Harvard Law School, 1947-52.
During World War II, Mr. Amory enlisted in the U.S. Army
as a private and served with the 258th Field Artillery, 1941-42;
533d Engineer Boat and Shore Regiment, 1942-45; and commanded an
amphibian engineer battalion and regiment in New Britain, New Guinea,
Luzon and Southern Philippine campaigns and in the occupation of
Japan. He was discharged as a colonel in 1946. From 1947 to 1952
he commanded the 126th Tank Battalion of the Massachusetts National
Guard, including active duty as a student in the Infantry School
and the Command and General Staff School, 1951. He is now a
colonel of infantry in the Organized Reserve Corps.
He is author (with R. M. Waterman) of Surf and Sand, (Cambridge,
Harvard Law School, 1947) and (with Covington Hardee) Materials on
Accounting, (Brooklyn, N.Y., Foundation Press, 1953).
Mr. Amory was assistant director in charge of Economic and
Geographic Intelligence, CIA, February 1952 to February 1953 and
has been Deputy for Intelligence since that date. He has also
served as CIA Adviser to the National Security Council Planning
Board since March 1953. He was a member of the U.S. delegation
to the Bermuda Conference, 1954, and the Bangkok Conference, 1953.
This is Mr. Amory's fifth lecture at The National War College.
Approved for Release: 2018/07/24 CO2437974
--ttfti(744,4;41
Approved for Release: 2018/07/24 CO2437974
INTELLIGENCE CONSIDERATIONS IN THE
- FORMULATION OF NATIONAL SLCURITY POLICY
By
The Honorable Robert Amory, Jr.
(21 October 1959)
COLONEL TWADDaL: (Introduced the speaker).
HR. AMORY: General Harrold, gentlemen:-
It is a pleasure to return to this platform to take up
this rather knotty subject with you.
You have had a previous session earlier in the year with
General Cabell and Lyman Kirkpatrick, our Inspector General, and I
understand that that covered the baCkground history and organization
of CIA and the various relationships of a managerial nature in the
Intelligence Community.
You have also had other speakers touching on or around my
topic. Cordon Gray showed me his talk to you of a week ago, and I
was pleased to see how much note he paid to the way in which intel-
ligence is absorbed and ground into the policy-making procedures.
So I brought no slides, no wiry diagrams with me today.
thought I would deal in as substantive way as possible with the
formulation of intelligence for the policy raker. I am going to
condense my 'remarks as much as possible - those of an expository
nature - in order to get, in the latter half of my talk, to a
series of problems with which we are still wrestling and: which you
might find of interest to pursue in your thesis writing and dis-
cussion croups.
- 1 -
--
Approved for Release: 2018/07/24 CO2437974
--4;L442,2141
Approved for Release: 2018/07/24 CO2437974
The basic national security policy papers as they are annually
revised contain an intelligence paragraph, a sort of mandate to the
Community, that has changed very little over recent years. This para-
graph lays a threefold leauireEtent of deve3oping and maintaining an
intelligence systua capable of collecting the requisite data on and
accurately evaluating three types of things:
1. Indicaions of hostile intentions that would give maxi�
mum prior warning of possible aggression or subversion in any area of
the world.
2. FstfzILT;cCcapabilities of foreign countries, friendly
and neutral as well as enemy, to undertake military, political, eco-
nomic, and subversive COUT3es of action affecting U. S. security. And,
3. Just to be sure we had not forgotten anything (1 was
one of the drafters of this paragraph), we put in a little itaa "e",
forecasts of potential foreign developments having a bearing on U. S.
national security.
Obviously, if anything was left out of the first two the
third was intended to encompass the waterfront.
Now, I will talk first, if I ray, about the advance yarning,
the first mission of giving this maxiL.um prior yarning to the policy
gakers. Of course, in really critical times it would te directly to
the action components the unified co=anders, the Joint Chiefs of
Staff, etc. There are three aspects to this.
One aspect is getting the intelligence :Tar infor,�ation in
fast from the field, in sufficiently short time so that in this hectic
_
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
age in which we live action can be taken on it in a timely fashion.
For this we have developed a thing .we call a "critic intelligence
system".
First, we have defined throughout the CoMnrunity on an agreed
basis certain types of information Which is characterized as being of
�
such a nature that it might have to be brought to the immediate at-
tention of the President. Please realize that in making any defi-
nition like this (and I won't bore you with the types) that an indi-
vidual person, whether he is a clandestine case Off-leer in the Middle
East or an MA operator in a station in Okinawa or Alaska, or something
like that, is not going to have the big picture. He has to be able
to judge something by one nugget at a time, one form, and ilo:1:c up his
mind whether that is an item that had better be fed back jolly fast.
Assuming that he sees something like that, he slugs it "critic" and
puts a flash or emergency precedence on it, depending on whether in
his judonent it is the type of thing that ought to get back in ton
minutes or in an hour.
That message then goes through with maximum clear trackage
in the communications network, and the important thing is that when
it gets to Washington no further human judg-lent is applied to the
question of whether or not it is disseminated throughout the Community.
If it has the word "critic" on it, bang, it goes out from message
centers by an automatic multlple teletype system in a matter of
seconds and then again it is automatic that the watch officers on
24-hour duty throughout the Intelligence Community immediately bring
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
that to the attention of their chiefs -- in our case to the attention
of General Goodpastor at the White House for the President.
The theory in the pnocoo is quite different Iron the facts
to date, because, as many or you who have been engaged in communi-
cations know, we are not yet really within striking distance of the
ten-minute goal at the mancnt. But thence is a program, which is
approved by the ESC, for implementation by the Department of Defense
with Guidance from the Intelligence Ca=anity that looks toward this
when ve get adequate automatic relay and routing systems electronically
controlled and run. But we have made a vast improvement and a few
figures I think might be of interest to ;you.
In the crisis or the fall of 1957 (maybe forgotten by many
of you in the light of the more dramatic crises of the ouraer of 1953
in the Hear East), the one in vhich Syria was thought to be going
Communist and the Turks and others were planning to make a quick
intervention and in turn the. Russians were threatening intervention
in Turkey, we had the kind of situation in which Washington should �
have been apprised as early as possible. Yet, the messages coming
back at that time averaged nine hours and 35 minutes from the time
of filing, generally in the Middle Eastern area and in sale cases in
Moscow, until receipt at the policy level in Washington.
The following summer, putting into effect the "critic sys-
tem" for the first time, a year ago July� a similar type situation
produced an average time of one hour and 23 minutes.
This year we ran a test in September, not having a good
- 4 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
crisis to contrast to that, which we really made a little rare dif-
ficult than a real one because of various control factors. We picked
various and sundry out-of-the-way-places in which intelligence infer-
ration might originate and tired a large caries of ressages right
straight through to the white House. There our average time was
down to 59 minutes. The Eediau tole, which is probably more ivpor-
tant because one or two erratic long-shots weighted the average up-
ward, was only 46 linutes. So we are beginning to get a syiteel that
is quite on the way to being effective. But in the days of 1/allistic
missiles and that hind of thing, we obviously cannot be satisfied
until we get further ahead with the automatic signalling system.
The second thing that you have to have is an on-duty group
at all times of the day and night and all days of the week, which we
now in the rational InCsications Center in the bottom of the
Pentagon. The absence of this, as many of you know who read the
history of the last week of iloveaber and the first week of DecelAber
in 1942., was really, more than anythIng else, to blame for the failure
of the national lc:sacra to appreciate what the Japs were going to do.
Everybody knew there was a crisis coming, but everybody's eyes were
on different parts of the world and on their ova fields of responsi-
bility; and the various bits and pieces were not rut together in front
of intelligence exerts in a single hand for observation and analysis.
This we think we have covered now in this Indications Center, particu-
larly in having professional intelligence officers on duty at all
tires of the day ex-xm fral all three Ser7ices, State Depi..1-'ment� and
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
CIA, so that there is never a case There a corporal or yoes:-:an is a
Lere CZ but that he is a person who has professional training and, is
ready to act; and the director or the on-duty man of the HIC can
call the Vetch 'Cormittee into being.
The Watch Comittee, as I think was explained to you, is
immediately subordinate to the United. States Intelligence Board. and
is chaired, by our Deputy 1)i5.-e.ctor, General Cabe11. It meets at the
drop of a bat; it has sometimes net on. nine or ten minutes' notice;
It meets right in the rational�Indic.:ations Center, :where all the
I'dopenthel'e on hand .or flowing in within the next few minutes will
be available to them.. They make up their considered juda).(.?.nt as to
whether or not this is an.indlcation of hostilities or a situation
that can be exploited in a hostile way by the Soviet 'bloc or any
other major force that thretatens our security.
Finally, of course, if there is time, there will be a meet-
ing of the U. S. Intelligence Board and a crash national estimate
(which I will discuss in more detail in a fc-cr minutes) readied. for
White House consunption. All this can be telescoped. �
I mean, if it is a really hot item and. a really urgent
Latter, the Watch Con:-,litte.o would. meet over the tele-ohone. Secondly,
the Naming 'from the ration Indications Center and fro-a General
Cabe11 Would. imediately co to the action authorities - the JCS and.
the White House - even as the evaluation of thendoper was taking
place. Finally, it is a prdLT.c that we have wrestled, with from time
to time and that is What I call the quee.3tion of the width of the
- 6 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
focus of the Watch Cormaittee.
If we nal-e this Watch Coadttee in the Ilational Indications
Center respo-:,sible for slerting the Govoralent to every little possi-
bilitnthat is going to be anneyi.1/4; and possibly embarrassing and
confusing, such as a rovolution or a E___Lm,o" in Bolivia or an assassi-
nation in Ceylon) or something like that; they are just going to be
so diffused that they are not going to be alert enough to pick out,
series of indications that ray be salothing of much more -1,r1Daztance
to us in the way of being related to major hostile aggression. There-
fore, we limit their responsibilities, as I say, either to hostile
actions directly cmerc.,-tecl. by the Soviet bloc or situations in or
near the Soviet bloc which aro'carable of ready and rapid exploitation
in a hostile way by it.
If there is any doubt in the case of a situation generating,
the Watch Cc,r2aittc,... consults J!th the USI3 and says: Do you wantus
to take this under advisement and watch it or don't you? The U3ID
will frequently say: Leave that alone. We will camission a special
task force to keep our eye on that. You watch for the big ball g9ne.
Finally, there is the question of being sure that everything
necessary gets to this group. It is all very well to sot it up, but
if highly sensitive infoli:ation is going to be withheld from it by
somebody who is afraid of leaks or trusts only people who have multiple
stars on their shoulders or ,325,000 a year civil pay salarics, you
won't get =yr.:here and, naturally again, vhich you had at the time
of Pearl Earbor.
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
So we got the rsc five years go to direct every Gover=aat.
departiaent and agency to make fully available to the Watch Committee
all infornation and intelligence of reasonable credibility pertinent
to its Mission and functions vithout restriction because of source,
policy, or ancrational sensitivity. (Of course, when I say "get the
NSC", as you well know, it Leans get the President on the advice of
/Inc to Give co :maid direction.)
This took so]e doing and it has not always been lived up
to, but its spirit is never attacked, and when peo1.1,e are caught out
on it they are contrite and therefore say they will try to do better
the next time.
In addition to that, yhiCa require intelligence to cane
before it, it is also neeesrary,. in order to judge accurately enc..T
reactions, that you have what we are up to at the same time in a
given crisis. So the second paragraph requires the Watch Committee
be kept informed concerning significant diplomatic�political, mili-
tary, or other course of action by the U. S. approved for immediate
implementation or in process of execution which might bring about
military reaction or really hostile action by the U.S.S.R.
I 'would like to pay tribute here to the military Services
and the Joint Chiefs as a corporate body for the vay in which they
have lived up to that in the off-shore islands crisis, and other
things like that where in order to judge what the Chinese were up
to you had to cran% into the machinery what the situation must have
looked like in Peining in view of the movements of the Seventh Fleet,
8 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
and DO on.
We think, thus, that we haTe devised a system that takes
full account of failures of 17 - 18 years ago. Of course, it can
still be defeated by her:Ian error or human incapacity. But we be-
lieve that as far as an abstract organization is concerned., and
putting the means at its disposal, the Community has done about
everything possible to see that intelligence vill not be ignored
in that kind of a crisis.
Of course, you do not want to confuse this with the early
warning system -- this we call the advance warning systea -- the
early warning systen of actual hostile moveuent of rissiles or air-
craft as detected by the various warning lines. The two are very
tightly integrated. Anything that came over the DEW Line, or coe,e-
thing like that, would be fed into this Watch rechanism, and, of
course, the Air Caamand mist and EORAD and others would be getting
any of the advance intelligence indications. Put this is a respond.-
bility that exists prior to getting things handled prior to the actual
mechanical detection of hostile action.
So much then for the first question of early warning.
Now, to the question of what we do about eatirating capa-
bilities and. treads and sort of forecasting developnents in aid of
the policy maker.
The machinery I am going to talk about, of course, deals
with what we call and define as national Intelligence, that which is
required for the fonallation of national security policy, concerns
- 9 -
�rirepi;07.2...
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
more than one depar'Q-sent or agency, and. transcends the exclusive
competence of a single de.partr.r,ent or agency. That is contrasted.
with departmental intelligence, ''-shich is that -which any department
or agency TO-Craire.3 to execute its own mission. Obviously, the two
overlap and. anybody lilac) thinks one can draw a line beteen them and
say �that the national estimates machinery %rill handle one to the
exclusion of, we'll say, 07..TI or ASTIA. (Any) has rocks in his head..
The iriDortant thing is that if the Chief of ravel Operations or
the Secretary of the llsory has need. for certain intelligence to be
developed in a ,say suitable for his planning and. programing he can
order it vithin his OIM resources. And. just because it is a naval
matter shall ye say, how ma.ny nuclear -submarines do the Soviets
have and. what is their progase,m for building them -- does not mean
. that it is not also national intelligence, because- if it is irsportant
enough to affect the national security then it can also be handled.
in the national rachinery. The only important thing is that once
it has been handled in the national machinery, with everybody having
the right to speak and to be heard on it (which I shall describe),
then that is the national intelligence on the subject until revised.'
and. it is not remissible and not sensible practice for a given
department or agency ronila.terally to put out conflicting estimates.
We have bad no trouble with that over the last several years.
We have been in business in this national intelligence busi-
ness now nearly a decade, and after the first two or three years its
general acceptance has been widespread. But there have been occasions,
- 10 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
not so much in the Services here in Washington as in commands out in
the field, where the commander will say: "We have a very interesting
paper out of Washington that is called National Intelligence Estimate
Number so-and-so. On the whole we think it is rather good but we dis-
agree with this, that, and the othei. thing, and the command will be
advised and guided accordingly." That also has not happened in the
lest couple of years.
Now, how do we get this national intelligence? How is it
done? I think it is worth a little chronological narrative run-
through.
In the first place, the statute would give the Director
complete authority to do this on his own with his own machinery. The
statute was intentially vague and broad, like all good organia statutes
in this and other directions. Dut if he aid that and to the extent
that it was done in the very early days of the agency, he, of course,
would get nowhere, because, when it is presented, no natter how well
it is dressed up and how gaudy a raiment you put around the term
"national intelligence estimate", if in the policy meetings at the
level of the NSC, the Cabinet, or otherwise, the Secretary of Defense
says, "that's fine, but that is not what we think over in the Penta-
gon", it obviously is no more than a fifth wheel in other departments'
and agencies' views.
Conversely, this could have been done on a pure committee
system. It could have been done by whacking out contributions -
stap/ing them together, and saying: Here is the Army view on the Army
Approved for for Release: 2018/07/24 CO2437974
-croz44.14.4_
Approved for Release: 2018/07/24 CO2437974
thing; here is the Ilavy view on the Navy side; the State Department
on politics; and so on and so forth; and then handing it over uith
everybody approving their own pertinent section of it. If you did
that, you might have agreement on everything but you would have no
corporate responsibility for the whole and you would have no integrity
in the paper. It just would not tand together as a single useful
docuocnt.
The compromise reached, whether you call it midway between
two possibilities or anynhere in between (at least, it is not one of
the two e,:tremeo), is essentially the committce system in the sense
of final corporate total rezTonsibility by the Board for the paper,
but a process of developing thd par.er that focuses responsibility
on a special rachinery which happens to be on CIA's payroll, but is
very definitely machine information, and on behalf of the intelli-
gence Community - to wit, our Office of I:rational Estimates.
The Office consists of two parts: mainly, a mall thing we
call the _Board or rational Estimates -- approximately ten senior
individuals with wide experience in Government and in intelligence
usually, drawn from all walkn of life, or at least from as many walks
of life as we can get ten people that are pertinent to the intelli-
gence game, including normally, but not necessarily always, an ex-
Service officer from each of the three Services, an ex-member of
the Department of State whether Foreign Service or, as in one case,
an ox-Assistant Secretary, a couple of historians, a lanycr who was
General Counsel of NSA, and so on and so forth.
- 12
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
This is a board -- you might call it agglomerate, or some-
thing li1e that -- that has two rajor features to it. One is its
total separation from adrani atrati ye reaponsibtlity. It has nothing
to do but do this thinking and estimating, cashing its ova individual
pay checks every fourteen days, or something like that. And the
second is, insofar as possible, having no responsibility or concern
with policy or programs or any loyalty to a particular Service or
department or national Ipolicy� it should be able to look an objectively
as huaanly- possible at the vorld events as they go by and portray --
no axe to grind, so to speak.
It is supported by a sniall staff of relatively younger uen
yho are drawn from largely academic circles initially but who have by
and large experience of six to eight years in the intelligence busi-
ness vho are coape tent draftsmen, good thorough readers of the world
scene, and, again, partake of this independence of position of their
seniors on the board.
You were assigned in your reading a short article by alMOB
Burnham. You will note in his highly critical observations on our
political and strategic intelligence, as opposed to what he calls
technical intelligence, that he calls for just such a group - a ,sva11
group - and he says: Leave then alone as far as intelligence is con-
cerned; just give them the Economist and the Neu Yorlz..11;_r:1(..
Well, that is not a bad idea sometimes, and I am sure they
all do read the Nell Yo1-71: Tires and. the Econoa!ist, but there are other
things that occasionally add a little useful grist to the mill. The
13 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
important thing is that we have tried to get just this kind of 'inde-
pendent mentality and take them away from problems of internal security
and administration of large attache oranizations or fpreiga service
problems: and policy. The only difference between what Durnham asserts
�
he would like and these people is that these people have an open
raind. and. Burnham would like them all to have a closed Lind.
row, hoJ do these papers get put together?
In the first place they have to be conliszioned or born.
On that Lr. Gray, I think, gave you some inclination of the close-
ness between workings of the Council planners executive secretariat,
himself, and those of us who are coLlai3o.i.oniAr.; NiEs. In fact, it is
a fairly free cor:ulission. If the Secretary of State calls up and
says he is worried about the situation In Yugoslavia, he can jolly
well have a national esthete on Yugoslavia; also, the Secretary of
Defense, likewise, or the Joint Chiefs, or anyone.
By and large we try to focus the efforts of this body, be-
cause it is limited in the number of estimates it can do well during
a given year, to the agenda of the Council. That we try to do in two
ways.
One is: We will provide an esthete for what the Council
plans to take up. But, conversely, and more importantly, being in
the intelligence business what we decide the Council ought to be
thinking about we will have a national intelligence estimate geared
up so as to stimulate and very likely obtain a review by the Planning
Board or the 003 or the MC itself on what is going on in a given
- 14 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
area.
Vore and Lo.fe, these :caper:, are responsive to real questions
that confront the T;olicy maker. In the early days they tended to be
very largely y'.oat we call "country studies" -- the outlook in France
over the next two years, or the outlook for Algeria, or sorlething
like that. Today, while we still'have a certain alount of those
papers in the accoint, We Give priority to payers that say: Uhat
will be the world. reaction - Carrunist bloc, neutrals, allied, and
so on - to the followng U. S. courses of actionwlth respect to the
Horn of Africa? or sooething like that.
The State Dcoartr.ent's Policy Planning staff, ISA, and the
Deparnent of Defenoe, or otheiolse, uoiking with army Lay and
Yr. Gray, will define the questions of the policy options in a
realistic way, and then we will Give olor best prediction as to what
will happen if one or the other or none of then are adopted. The
nost recent jutty job done like that was last spring on certain
courses of action to contend with the Berlin situation, including
opening up a corridor, and co on and so forth. I think Yc. Gray
discassed with you one of tho.:A with respect to the niddle East.
The next stage Is the one of getting the ter_s of reference
of the estimate reasonably defined and detailed to the end that if
intelligonee can be found to answer the questions you will have been
sure that you asked all the questims that the policy naka'would
like to have answered. Then these are yhacked up ti.:;on.,3 the various
agencies and contributions solicited snd obtained fron the estl-aates
- 15 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
sections.
The important point to stress here is that the contribution
that comes in iG a genuine study of all the aspects of that problem
and is not a set draft of how the submitting ageney would like to see
�
that part of the esthate read. For that reason the contribution
will norJally exceed the final estimative section 7hich it subtends
by a ratio of ten to one in rages or words.
than all these contributions are in, a small task force of
this junior staff that I talked about gets to work and whacks out an
initial draft, staying within the limits of the contributions in
almost every case. In other words, they are not supposed to have
independent ideas that are outside the right or left field foul line
of what somebody in the Community thinks about the situation. But
we do not make that an absolute grettad rule, and from time to time
we will startle the Cammlity. by coming in with our initial draft,
taking a position.Lore extreme on a given point than any one of the
responsible contributois has taken.
When that draft is completed it is sat on by this senior
board that I described as a"Llurder board:, refined, beaten around,
redrafted, and then resubmitted to the contributors, the meshber
agencies of the U3IB. This is called then the "Eoard Draft". They
give, if there is time, the agencies time to think it over, to come
in with their recommended. changes, Which are froluently lengthy and
widespread. They come back and meet with the panel of the board in
the final or semifinal heat of the tourna:dent. In the CbUrSC of
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
their work, the dissents are nornally for the first time hammered
out and set forth. When their work is done, the paper goes to the
United States Intelligence Board for its final working over and
approval.
Now, I stress here, as I have almnys, that the action of
the United States Intelligence Board - the heads of the intelli-
Gence agencies - is by no nonns rubber-staming or cursory. They
take the attitude that in this process they are not just the senior
representatives of their intelligence services concerled, that they
are a corporate board of finnl evaluators of intelligence infor-
vation, and they stake their sort of personal as well as &wart-
rental reputations on having this paper as gooa as humanly possible.
when that process is through you have a National Intelli-
gence Estimate.
There are two or three things to say shout theol. For
those of you who read than or commisslon than, as you all do in one
way or another or have done before, the first is the ioportant
requirement of reacling them with cave. Little adverbs, little adjec-
tives "it is bearly possible", "the chances are slightly more
th,on even" -- are essentially matho-latically thought out before they
go into the language of the paper.
You may say: Well, wouldn't it be better to quote odds in
a numerical way? There are .some of us who think it you'd not be a
bad idea from time to time, though others at the noment (well, the
rajority of controlling positions) say that would give a false
r 17
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
impression of accuracy and clarity that we do not have.
The second is to pay very close attention to the dissents.
\To encourage disJents on all matters of real substance. We try to
discourage th.6:1 when it is just a question of saying the sane thing
in two different ways.
By and large) nine-tenths of the 'dissents) you will find)
have a real matter of substance at str.n'ze between the footnote and
the text or whether evenly balanced between the two split texts.
These serve the purpes:." of calling the policy rilaker's attention (1)
to "the .fact that there is something �important at stake and (2) to
the fact that one or more of the agencies feels quite differently
from the body. What the policy maker does with that) of course), is
his own business. He may say: .1 am going to sit as an mpire�on
these and I will side with the majority or the minority. If he is
a wise planner) however) he will take into account the fact that the
minority is a thoughtful one and his plans) if humanly pow:Able)
should take into account the contingency or the point of view set
forth in the minority. At least) we avoid. this 1,Tay what we are
often accused in the public press of doing) least comon denominator
writing) tent language, which everjbody can comfortably get under.
with no clear direction as to what we mean or guidance to the policy
ra.1,1:er.
Probably that process has sounded to you extremely lengthy
and eimbersome. You can say: Well) no wonder we are so Tar behind
in various aspe.cts of the cold war) if that is the way everything
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
has to be ground through before somebody does something about it.
But the fact is) withortnissing any really essential ingredients,
this whole irocess con and often is telescoped into a day or two
or) in two or three certain instances in recent years, into a ratter
of a few hours. The contribution, instead of being a written one)
WILL be al_eeing at which the various Service and agency representa-
tives core over to our building and. give their views orally; some-
body jots down notes; souebody writes the first part of a draft and
hicks it back in; but essentially the intellectual process is in
that sane orderly manner. I recall one time when the U5113 was
actually sitting on the first part of a paper while the reresenta-
tives were arguing out another section of it. So it does lend itself
to rapid action. �
So many of the things that the policy maker really deals
with at the rsc level do not require that kind of rash and crash
action. Those are more apt to be On decisions on whether or not to
cancel a prograu or to put more funds behind this, that, or the other
thing in a given country. But when you are planning generally what
you are going to do over the next few years in Tropical Africa or
what your attitude should be toward Sino-Soviet differences, you can
take time to do a thorough job. And on a reasonable schedule, as we
have on our big estimates of the Sino-Soviet bloc
00 NO
a big Eussian
,paper, a big China paper, a big satellite raper, and $0 on and so
forth -- there is no reason not to Take those papers as good as pos-
sible by having a-maximum period of detailed study in detailed cross
- 19 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
fertilization of ideas in the process. So, as I say, the process can
take anyc.here frai a few hours to a few -.onths, and that is, I really
believe, as it should be.
We also include in this a post-rortol of the -whole pro-
cess Which is a very good aid to the collector. We will state in
that that this estimate was herPered by lack of information on this,
that, or the other thing, by conflicting infomation on this, or so
on and so iron, so that the collecting agencies in general terms
can go out and do a better job prior to the next cue.
We alo have happy as we are at the quality of our Board
and the representatives froi the agencies, a strong feeling that we
can use outsiders to great advantage. We have a body that mots at
Princeton which we call 'The Princeton Consultants" of really 'first-
class people -- you night c.?y the hind of people we would, love to
hire fur the Board, of rational Est3rates, but they are too happy
doing their other chosen occupations. To give you expLples, Colonel
Abe Lincoln of the faculty of West Point, George Kennan, Kanliton
Fish Aristrong, Bob Boyd, !lax Milliken, and so on. They mcetdith
vs about six ties a year for a day-and-a-half span.
Vaturally, we do not burden these people with everything
that is going on in the estinating world. We pick the hey ele.:ents.
They ray deal with an undeveloped area. They are nore apt to deal
with big questions of the Grand Alliance or, particularly, the Soviet
bloc. We used to Use then purely as post-mortemers. We would wait
until the paper %o.s done cad then sul-ait it to then get their comments,
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
discuss it with them, come back and sort of tell the Board where we
think maybe we vent wrong -- if we had had the enlightened views of
George Kerman on this we might have done so-and-so. But they got
bored with that and we decided that that was a pretty futile way to
do it.
So
now we tend to take papers to then at the stage of the
Board of Estate s draft. When it has been thought over by us) it
is in more than a raw state but it nonetheless has not gone through
the final interagency-polishing. Then we get their views which can
be cranked in and be helpful.
I do not think I have to take. very much time, in view of
Gordon Gray's lecture, to describe to you the role of the. Director
on the Council and the way these papers cone in as intelligence �
annexes and are either used in tote as such or are taken and con-
��������� ma*.a.
densed into the general con:Aderations rart of a policy peer.
It is my job at the Planning Board level to see that any-
thing which pertains to intelligence in an USC paper is precisely
reflective of the Co:-,-4tuitty' views. That is why, naturally, if
possible, you have an ND. If you do not have an NIL, than by scout,-
ing around informally I try to zee that that is accomplished, and,
at the highest level, the Director does the same thing.
There has been a little confusion, I gather from some in-
formants (not breaching the usml security rules of this College),
who told me that I either misspoke or was misquoted a year or so ago
about voting on the rsc. The fact is, of course, az ire Gray says,
nolody votes at the 170C or at the Planning Board. That is a corforate
- 21 -
'.Thrrel..9.1
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
body of advisers vhe talk to the President, and when he has heard
enough talk on a subject he issues his executive action, as he con-
stitutionallymust, on his on soul responsibility.
The point that I vas trying to make is that the Director
at his level and I at my level, necessarily for our oun (and I ao
not mean personal) institutional good, stay out of the argument
under the heading "nut to do about it". We will keep them in
argnment on the point This is the Situation; se it clearly; see
it realistically, no rosy glasses, and so on and so forth. But then
Vhen the question comes, what to do about it, we limit ourselves to
predicting certain consequences if a certain line of action is taken.
That, of course, by no means keeps you shushed, because you can do
a fairly neat Italian job on a policy by making predictions about
what dire consequences to our grandchildren it nil have lf anything
so bizarre is adopted. But leave the adjectives off the policy;
just put thm on the conseuences.
At any rate, it is quite clear that we have to be careful,
if we are going to be the zlerchants of an intelligence there that
will, have accoptance,not to take sides on how much should we spend
on a given pnogra-ii or vhother or not to join n pact, or something
like that. I think it has been wisely our policy and is generally
accepted.
The important thing that I an trying to .get across in all.
this machinery, the relationship, is that you should solve insofar
as Inu:17,nnbeings can do It the intelligence questions in a foam
- 2,73 -
--lt,34.14Z�_
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
that is free from policy considerations before you enter the policy
forwn, and then the -policy Lakers should. be sufficiently dieciplin-A.
(1 will cone to this at the very end.) - not always so - so that they
will not substitute their olin instinctive intelligence judgfilents for
this catholic process judgclent that is presented to them.
Of course, nuch nore than riE5 go 111) to the Council and go
to the senior policy rakers.
We have a daily bulletin and we have a weekly briefing by
the Director of the ITSC. That has been an institution ever since
General Eisenhower bec=e President, and a very salutary one. It
gives him a captive audience.
You know *perfectly well (no reflection on our good friends
here in the audience frou the Treasury) that the Secretary of the
Treasury, has a lot of things to worry about other than the revo-
lution upcoring in the Caleroons and when he gets a daily bulletin
or briefing sheet on his desk in the vornIng the likelihood of his
reading that with care is not very great. But when he is sitting
around the table with the President, who is attentive to the Director
of Central Intelligence, listening to a briefing on the world situ-
ation, there is nothing nuch else for him to do except listen to it.
So you get a chance to have the high-level people ell hoisted aboard
on the key things.
It is a terrific *responsibility to figure out how to 1140
that fifteen rinutes of time, and a great deal of care goes into
that in our agency. Roughly speaking, a third of it will be targeted
Oa
23-
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
on the topic on the agenda of a policy nature, but the other two-
thirds of it will be the combination of the recent events of the
week since the last meeting as seen through intelligence eyes, plus
an occasional long-range roundup, such as: What is the meaning of
the failure (or not so much failure.- whatever it was) of the
Chinese coLnyne move4ent. There is no one week in which that is
particularly topical. But every so often the Director will pick up
or round up three months in the past and project three months in the
future some topic like that.
The daily bulletin, which goes to a selective group of
about forty of the highest level people and out by telegraph to the
major unified commands, used to be an unilateral CIA publication,
just because time did not seem to serve to coordinate it in anything
along the lines of an NIE. But, in fact)we have tried, on direction
of the President's Board, to make it a Community paper for the last
year-and-a-half, and have been quite successful. It has a certain
ground rule on it, though. We coordinate like hell on it from about
Lour o'clock in the afternoon until about 5:15) and then the argu-
ment ceases right then and there. The CIA and majority view pre-
vails and a footnote is taken by anybody who feels strongly enough
about it. Very rarely is this necessary because we are not making
predictions of the future in the bulletin; we are spotting various
intelligence items.
How good are these estimates?
I am sure I will get that question from the floor, so I
- 24 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
might as well get it now while I am fresh from myself.
By and large) on the biggest isaues) they have been very
good. We have not really goofed on the basic direction of where
Russia is going. We have not goofed on the general strength of
Communist China when other people were saying that the good Chinese
people would never accept Communism) and so on and so forth. We were
discouraging and not downcast about it, but were realistic about it.
We were good on such critical issues as how far would the Chinese
Communists go when theyopened up on the off-shore islands and at what
point they would. stop. We have been pretty good) we think) on what
is back of Berlin and how far they would go on that. And the re-
turns are obviously not all in on that.
But we have had very conspicuous failures. There is no
question but that we did not call attention in any adequate degree
to the degree of unrest in the satellites that produced Budapest
in 1956) and so on. In smaller ar...ms of difficult problems I think
our record is much better than average. Failures lay in the Algerian
situation primarily. We have been too pessimistic about some coun-
tries. Ambassador Chapin's country) I think, we sold down the river
a little more frequently than we should have. But by and large) we
have called attention to trends in these countries in a timely
fashion and the papers stand up remarkably well.
On the big final question) about which you see most in the
press -- overestimating and underestimating the Soviets' military
capabilities -- we all know the returns are not all in on that yet.
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
We have over the years (and I do not go back all the way to the days
before the true VIE rachincry on when they get the atomic bomb,
and so or, but taking the period fmil 1951, when General Smith caLe,
through to date) been like a line fairly steady with developments of
Soviet strength scattered around it in a scattergram fashion that
is not too discredible. But it can be very expensive when we are
off in a particular one, ae we were on the aircraft in 1956.
There we are blamed a little rore than we Should be because
we did not purport to say when they would have them- We ca-id.: "They
have a BISON and a ITATi in production and they are apparently flyable
and combat worthy, if they want to." The trouble is, we (Bd not
underscore the "if they rant te. Put the ray the thing was stated
meant: If they want to they have the Boeing plants, the Lockheed
plants, the Martin plants equivalent to produce these at a rate
scaling up along the line that will Give thcell Y-hundred of these
things in 1959. That was taken too nuch by too nany people as a firm
prediction that they would. Because the Russian was an evil beast
and we stood in his way, obviously-he was going to get anything that
would clobber us. Therefore, people just rapidly read over that and
said that a production capability was a prediction of what you'd be
in the offing. So that is a hit or miss In a way, but I think I have
hit the high points of the type of question.
When we post-mortem a paper we also post-mortem the old
paper and look at it in the light of the new one. They read remarkably
satisfactory in most instances.
- 26
Approved for Release: 2018/07/24 CO2437974
---e6saan ril
Approved for Release: 2018/07/24 CO2437974
Now, in the rcmainin:; time (and there isn't very Ifuch) I
would like to just tick off a few problems that are not really in
this area but iapinge directly upon the area of how we sv;ply intel-
licence to the policyiers.
In the first place, 1 :nuld just like to observe that I an
concerned (and this is persol-,a1) with the relative military imbalance
in the USIB. There are on it ten mc:bers, of 'whom four are normally
civilian end six are military -- which does not sound bad, but two
of the four are really only there in most instances in a nonentity
capacity, the representatives of the MI and the AEC. They are very
fine when subjects pertinent to internal security or atomic energy
are before vs, but they do not porticipate on other items, -which
comprise v,aybe 93 - 99 percent of the business. That leaves us with
a six-to-two ratio -- all three military Services, plus NSA., plus
Secretary of Defense officer, plus the joint staff. And, consider-
ing the number of issues that are before us and the whole country and
the policy la.21ers of a nonmilitary nature, I an a little concerned
at this balnnce; though we all know that the State Depart'aeat is not
incapable or keeping up its end, it has a very fine naval career
officer sitting for it.
Considering also the fact that.my boss, as chair_Lan, while
he is not inhibitive from having his constrong views, necessarily in
order to bring about a 1-odest concensus of opinion is not too forward
in pushJng extreme views himself but is, rather, a corporate chairc;.-ui.
It alFost uorhs C_own to a six-to-one military to political-economic.
�27�
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
Oddly enough, 010 of the solutions surclosted to correct this
imbalance a little bit is to have the ICA and USIA resented on the
Board. USIA, by the way, has a highly competent, though snail, intel-
ligence organiffation of its own; and ICA is, after all, many times
the best informed agency of the Government on an economic and some-
times economic-militaly situation in a given country. The Board is
not op7Josed $o Luck by the Llilltary, but it is just absolutely kicked
down "dead finish" by our good fric-Os from the Department of State.
So if they want this lonely splendor, why they have rot it.
Flow, a little word or two on sources of information.
The proble:0.1 after all, of intelligence esthates is very
simply stated in that they cannot ba any better than human ju r7laent
can make them on the basis of the infoniation that is ground into
them, and that means how good and whence cometh our intelligence
information.
As to where it comes from in the EIEs, that is almost In-
Possible to state because we are not dealing with spot sources or
spot items, But the mrs essentially rest back in the defartments
or agencies on their OW11 studies, the UIS type of thing, and so on
back. Ve can take our little bulletin, which deals with spot items,
and give you some idea of the general breakdown of our reports.
Let's take fiscal year l9r)-9 and the bulletin items that wore put
out.
Six bulletins a reek, averaging about seven items, $o
forty t:lmes fifty -- 2,000 items, roughly speaking, State cables
23 -
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
and elspatchc....s
percent.
ESA materials -- 17 percent.
Foreign broadcasting inforration reports -- 13 percent.
World press- services -- $ percent.
. Military attache reports -- 4 percent.
(The latter ray seem to you unjust; an erroneous figure.
It is not. It is related only to this bulletin item because by and
large nuch valuable intelligenbe estimates - state of the Ay or
the Navy or the Air Force of a given country - is not a spot item
that you report to the White House and the Secretary of State the
next day. In other words, this would be quite unrepresentative of
the degree of contribution to This that the overt reporting by the
Military MakeS. )
It is interesting to see that happens then we break this
down only among the items referring to the Sine-Soviet bloc.
The State cables proportion and dispatches is cut more
than half -- to 17 percent.
ESA rises slightly
5
FBI press raterials rore than doubles ---to over 30 percent,
Vire*
to over 20 percent
thus showing, regrettably, in how far we are dependent upon what they
in their own sweet time decide to tell us over the airwaves as far
as rajor events lack of their borders are concernel.
- 29 -
-"7:311ei.
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
I hesitated a little to throw those figures into this talk
because they can be quite misleading and, of course, they totally
ignore the quality or iportanco of a given it=. Obviously, one
report could be worth a thousand nedium
level operational typo reports; or one first-class .military
photograph of a new plane could be worth all the
or low �
attache
rumors
that we perhaps could get abou'z, there being a ncIr plane.
The . ortant thinz,to roz�e; ler about clandestine collec-
tion, -which is our business, is that it is L7ff1cu1t be-
hind the Iron Curta,in. I:a do not like to be crybabies and play
excuses, but we are u-7) ag_vinst the toughest, nost efficient security
the world has yet hrown (it inkes 1Iitle Schutpolizei, and so on
and so forth, look ,Just sick by co):1Darison), lacked by treendously
loyal or frightened InDulations. I hate to say it, but I an forced
to believe that you should underscore the "loyal" rather than the
"frightened". They are just plain not in the 12ood to be recruited
for any of this business behind it. There are plenty exiles and
emigres, and so on and so forth.
On the other hand, cl:_ndesLine intelligence can operate
in
areas where Security is weak, both against the bloc and against the
country itself, with a great deal of efficiency. I an impressed,
for exam:rae, right now 14-ith how rapidly we have gotten a pretty good
look at
-30-
Approved for Release: 2018/07/24 CO2437974
� --(41-mfa4:=1
Approved for Release: 2018/07/24 CO2437974
That is not quite horse's nouth stuff,
but it is only one or two renoved from it, and it is very helpful.
It is important, in view of the difficulty of getting
information from behind the Iron Curtain, I think, that everybody
interested in intelligence be 5tron313r
There nay be other reasons for saying 'let's not go too far too fast
with it", but strictly fror where I sit and. where those in
intelli-
gence sit what the Soviets can learn about us by having a lot of
travelers, whether they be scientists, nilitarn or other wise (I
could not c':xe less), colTared to that we vill learn if we strike
�
an even and tough b:,rgain on the nature of the trip, is such that
we vust be careful not to let benighted Ideas or intorn%1 security
deprive us of a real net advantage that we win get in that -my.
Finally, before my closing, is the question of relation
with allies.
I do think that we have to be awfully careful in this busi-
ness to get the old rubber stamp "IToForn" firoly under centre'. Here
we are, with the defense of the United States absoluely integrated
with that of Canaal, or at NOBADI with a Deputy who is General Partridge
(or if Pa: ridge is :-Tray, why Air Marshal vho is actually
the man who Is going to defend our rives and children if a blow cones
tonight. Yet, when so-lebody gives re some intelligence and says,
"I can't pass it to UORAD" or "I can pass it to OPAD on a 'US eyes
only' tariis1; that is 0-1e hell of a note. And 'we have ce-laints fro:,
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
NATO the scare way.
We have excellent relationships with our British Common-
. wealth frienda. By and. large, as you know,
But that is
not enough to satisfy our friends in Germany and elsewhere.
At the same time, I must admit, it is not an easy problem,
because, when you get two Communists in the Icelandic Cabinet or a
totally fudgy security situation in Turkey, or something like th^t1
free4leeling with very sensitive data, NATO is obviously impossible.
But our tendency then in doubt, seems to me, should. be to be forth-
coming. We are running a grand coalitiw and we had better not
think of ourselves, strong as we are pro7Jortionately to them, as in
a position to withhold from them.
�
Finally, and. most :important -- and. this comes up all the
time in my discussions with Service people and occasionally with
State people -- remember that differences on policy about a given
country are reasons for rather than against exchange of hard intel-
ligence on that Country. An obvious example is China. The Britisth.
recognize it; we do not. But that is just the more reason that we
try to be jointly and severally as well informed. on China as possible
so that we can profit from what they know and. they can profit from
what we know and. let the policy makers have their arguments but on
a common basis of fact.
Finally, let me say on this business about getting consumers'
to use intelligence correctly that it is always a trial. For some
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
reason or another, there is a tendency on the part of all of us to
be our own intelligence officers. I was certainly guilty of it when
I was a Junior field officer of a co-and in the var. In fact, vly
intelligence officer was nore frequently than not attached tory S-4
to Go bael: to Army rear to find out why we were not Getting our
share of the varn beer becauee Ivas perfectly sure I knew -what was
up in front of me. So I an no exception to that. But it is anaeing
how people, who will rot second-guess their medical officer on the
amount of penicillin they need or their J-3 even on the anount of
training time reauired, will have their own ideas on intelligence.
Secondly, there is this pro"ole.. which I charac'eerieed once
before in this roan as the "succulent taste of the raw poop". There
�
is nothing that the boys at high level like better than a nice little
flimsy sheet of paper that has three or four Garbles in it, and so
on and so forth. And once they have wasted ten minutes pondering
that, when the intelligence prinlysts have cone up with a corrected
copy and an analytical statent about it, they are too busy: "1
an sore; "I know all about that". So you have to do something
which is very difficult: -withhold that from your boss. And then on
the white for.a: "Gee, did you see this?" He hasn't seen it because
you have been sitting on it, waiting until you get an interpretation
from him, and you have virtually Got yourself a heave ho.
, Finally, there is the question, as I say, in this same thing,
or getting people to use intelligence correctly, of getting over
prejudices and e-otioral involve=ent in the cold war. Uo"-.;.ocly waits
��� 33 -
......g177,5g1ZZ_ �
Approved for Release: 2018/07/24 CO2437974
Approved for Release: 2018/07/24 CO2437974
to be less dedicated to the general favorable outcome in that war
than I do. But if you are in the intelligence business and if you
are a good consuaer of intelligence, you have to be careful to see
facts hard and. clear and not batten onto favorable little whisps of
info=ation and discard major unfavorable ones just because you
would like to be On C. winning, winning, driving side all the time.
A typical recent example of that is Tibet.
1 reret to say that the probable (note that 1 say "probable")
true facts (nobody really knows) are that basicall
3'
we go. But 1.am perfectly willing to have us do that in the United
Vations, and so on, so long as we do not kid ourselves and think that
all two-million Tibetans are just Cheering for us all the tine, when
in fact they are not.
Well, that is kind of a rambly windup to a raMbly talk.
Would just like to say that I am well aware that we need
a lot of iupiovement in the intelligence picture. We are working on
it constantly from the opposite point of view from that of coplacency.
By and large, when people complain right out and say intailiGenec IS
not doing its stuff, I would only remind you that insofar as I can
�
make out from a career in and out of Government in the military the
policy 3;,1k.,-xlner today has more information with less II;ar3in of error
- 311.
Approved for Release: 2018/07/24 CO2437974
-
Approved for Release: 2018/07/24 CO2437974
than the average businessman has in making his decisions or the average
riaitary man has in time of colrbat.
Thank you.
- 35 -
Approved for for Release: 2018/07/24 CO2437974