STUDIES IN INTELLIGENCE. VOL. 1

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78T03194A000100010001-2
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 144
• Document Creation Date: December 20, 2016
• Document Release Date: May 5, 2005
• Sequence Number: 1
• Publication Date: December 9, 2003
• Content Type: SINT

For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
":`.; is{(.a A!iLh vC.J & ittCUROS (L1 Il
1,04MiA RLY AMR USE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78T03194A000100010001-2
FOR OFFICIAL USE ONLY
The Need for an
Intelligence
Literature
Articles by
Sherman Kent
and the Editors
I
STUDIES IN
Intelligence
CENTRAL INTELLIGENCE AGENCY
OFFICE OF TRAINING ? SEPTEMBER 1955
N
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
FOR OFFICIAL USE ONLY ' FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
All opinions expressed in the Studies are those of the
authors. They do not represent the official views of the
Central Intelligence Agency or of the Office of Training.
Address all comments and inquiries to ...
Office of Training ? ext. 2428
Subsequent issues will be disseminated widely throughout the Agency!
To make sure of receiving copies, or to secure extra copies, please
call the extension listed above.
NTRODUCTION
)y The Director of Training
l N the two articles that follow, Sherman Kent and The Edi-
tors explain in detail why we are starting this mono-
graph series, Studies in Intelligence, and how we are
going about the job. I should like first to explain the posi-
tion of the Office of Training.
Our chief responsibility here in Training is, of course, to
teach. In order to do this effectively, we have to be very
clear in our minds that what we teach corresponds to the best
intelligence doctrine and method available. To train in-
telligence officers, we must first define and clarify those basic
principles which should be the common property of every
member of the profession. The basic principles I have in
mind concern the mission of intelligence, and the organiza-
tion, techniques, and methods of intelligence activities.
From the beginning, the Office of Training has necessarily
been engaged in the production of written materials to be
used in specific training courses. But it has become increas-
ingly apparent to us that there is a need for a broader
approach to the problem of building up a foundation of doc-
trine and method which is basic to the intelligence profession
and all its activities. It is also clear that we have a degree
of responsibility for meeting this problem.
I believe that the production of these Studies will be a step
in the direction of creating a literature of basic doctrine and
methodology useful both to the training activity and to the
MORI/HRP THIS
PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
FOR OFFICIAL USE ONLY
Agency as a whole. In sponsoring this endeavor, I therefore
urge your active participation and support so that we may
all benefit in advancing the profession of intelligence by this
means.
MATTHEW BAIRD
Director of Training
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
THE NEED FOR AN
INTELLIGENCE LITERATURE
by Sherman Kent
I N most respects the intelligence calling has come of age.
What has happened to it in the last fourteen years is
extraordinary. Maybe our present high is not so ex-
traordinary as our low of 1941. In that day the totality of
government's intelligence resources was trifling. We knew
almost nothing about the tens of thousands of things we were
going to have to learn about in a hurry. As emergencies
developed we round ourselves all too reliant upon British
intelligence. Many of us recall important studies issued by
US intelligence organizations which were little more than
verbatim transcripts of the British ISIS reports.
In 1941, the number of people who had had prior intelli-
gence experience and who at the same time were available
for new government assignments in intelligence was very
small. There were few in Washington who could give any
guidance as to how to go about the business in hand. What
intelligence techniques there were, ready and available, were
in their infancy. Intelligence was to us at that period really
nothing in itself; it was, at best, the sum of what we, from
our outside experience, could contribute to a job to be done.
it did not have the attributes of a profession or a discipline
or a calling. Today things are quite different.
Let me briefly note the principal assets of today's intelli-
gence community. To begin with, we are at strength. Per-
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
2 FOR OFFICIAL USE ONLY
haps we are not as strong as the present volume of work
requires, but by and large we have the staff to do the man-
sized job before us.
Again, we are not novices at our business; we have a lot
of experience behind us. We are officered and manned by
a large number of people with more than a decade of con-
tinuous experience in intelligence, and who regard it as a
career to be followed to retirement. By now we have order-
ly file rooms of our findings going back to the war, and we
have methods of improving the usefulness of such flies. We
have orderly and standardized ways of doing things. We
do most things the right way almost automatically. We have
developed a host of new and powerful overt and covert
techniques which have increased the number of things we can
and do find out about. Most important of all, we have with-
in us a feeling of common enterprise, and a good sense of
With these assets, material and experiential, intelligence
is more than an occupation, more than a livelihood, more
than just another phase of government work. Intelligence
has become, in our own recent memory, an exacting, highly
skilled profession, and an honorable one. Before you can
enter this profession you must prove yourself possessed of
native talent and you must bring to it some fairly rigorous
pre-training. Our profession like older ones has its own
rigid entrance requirements and, like others, offers areas of
eneral competence and areas of very intense specialization.
g
People work at it until they are numb, because they love it,
because it is their life, and because the rewards are the re-
wards of professional accomplishment.
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
FOR OFFICIAL USE ONLY 3
Intelligence today is not merely a profession, but like most
professions it has taken on the aspects of a discipline: it has
developed a recognized methodology; it has developed a
vocabulary; it has developed a body of theory and doctrine;
it has elaborate and refined techniques. it now has a large
professional following. What it lacks is a literature. From
my point of view this is a matter of greatest importance.
As long as this discipline lacks a literature, its method,
its vocabulary, its body of doctrine, and even its fundamental
theory run the risk of never reaching full maturity. I will
not say that you cannot have a discipline without a literature,
but I will assert that you are unlikely to have a robust and
growing discipline without one.
Let me be clear about this literature that we lack. First,
let me say what I do not mean that we are lacking. I do
not mean the substantive findings of intelligence. Manifest-
ly, I do not mean those thousands of words we disseminate
each day about past, present, and probable future goings
on all over the world. I do not refer to the end product of
all of our labors. We produce a great deal of this sort of
literature and possibly we produce too much of it. It is not
that literature that I am talking about. What I am talking
about is a literature dedicated to the analysis of our many-
sided calling, and produced by its most knowledgeable dev-
at ?ec The sort of literature I am talking about is of the
nature of house organ literature, but much more. You might
call it the institutional mind and memory of our discipline.
When such a literature is produced, it does many things to
advance the task.
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
4 FOR OFFICIAL USE ONLY
The most important service that such a literature performs
is the permanent recording of our new ideas and experiences.
When we record we not only make possible easier and wider
communication of thought, we also take a rudimentary step
towards making our findings cumulative. We create a stock
of relatively imperishable thinking which one man can ab-
sorb without coming into personal contact with its originator
and against which he can weigh and measure his own origi-
nal ideas. His large or small addition to the stock enriches
it. The point is reached where an individual mind, capable
of using the stock, can in a day encompass the accumulated
wisdom of man-decades of reflection and action.
Consider such disciplines as chemistry or medicine or eco-
nomics and ask yourself where they would be today if their
master practitioners had committed no more to paper than
ours. Where would we be if each new conscript to medicine
had to start from scratch with no more to guide him than the
advice of fellow doctors and his own experience? Where
would we be in medicine if there was nothing to read and
nothing to study, no text books, no monographs, no special-
ized journals, no photographs, no charts, no illustrations, no
association meetings with papers read and discussed and
circulated in written form? Where would we be if no one
aspired to the honor of publishing an original thought or con-
cept or discovery in the trade journals of his profession? It
is not impossible that blood letting would still be considered
a valuable panacea and exposure to night swamp air the
specific for syphilis.
The point is that in the last few centuries we have accumu-
lated an enormous amount of knowledge. And the fact
that this accumulation has taken place since the discovery of
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
printing from movable type is by no means merely coinci-
dental. The translation of new thought into words, and the
commission of words to the permanence of print, more than
anything else has made possible a progressive and orderly
advance in all disciplines and all areas of learning.
In our calling, I am saying, we do not do enough of it.
To be sure we do do some writing. We have produced a
good many Training Manuals of one sort or another. We
have done a good bit of chronicling of interesting case studies
with an educational end in view. We have made transcripts
of oral presentations at training centers. If you ransacked
the "libraries" of intelligence schools you would find quite
an amount of written material. Even so there is a very con-
siderable difference between this volume of written material
and the systematic professional literature I am talking about.
It is hard to define such a literature, and I will not try to
do it in a sentence or two. As a starter I will note what I
think to be three important aspects of it. To begin with,
the literature I have in mind will deal with first principles.
A portion of it will 'certainly have to deal with the funda-
mental problem of what we are trying to do. What is our
mission? And as soon as that question is submitted to care-
ful analysis, there is no telling what will emerge. One thing
I think is certain: that is, that we have many more than a
single mission and that many of us have been confused not
only about the number and character of the many missions,
but also how each of the many relates to the others.
Another first principle that will have to be elaborated is
how we are going about our mission - what is our method?
Here again we will find out, when the question is systemat-
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
6 FOR OFFICIAL USE ONLY
ically answered, that there is not a single method, but that
there are dozens of methods; and from further examination
or discussion we will confront a good many new concepts
which will speed our task and enrich our product.
Let no one feel either that we are necessarily sure of the
nature of our first principles or that dispassionate examina-
tion of them would be a waste of time. In recent months the
intelligence community has had to wrestle with such funda-
mental concepts as "national intelligence objectives" and the
criteria for the selection of such objectives; the nature of
"warning"; the role of "indications" and so on. The results
of these discussions have been generally praiseworthy, but
the amount of time consumed and the consequent delay of
important decisions quite otherwise. An analogous situa-
tion might be a consultation of surgeons deadlocked on a
discussion of the nature of blood, preliminary to handling
the emergency case presently on the operating table.
This takes me to a second thing which I would expect from
a systematic literature of intelligence: a definition of terms.
Hastily let me add that I am not proposing that we write a
dictionary. Words which stand for complicated concepts
cannot be defined by a dictionary. Words like "liberalism"
and "democracy" require the equivalent of scores of diction-
aries, or scores of shelves of dictionaries. You cannot define
those as you define "paper" and "ink." So with our own
words that stand for complicated concepts - such as "evalua-
tion," "indicator," "capability," "estimates," and so on. As
of today we use these words easily and often - yet one won=
ders if they are always understood in exactly the way in-
tended. For example, we would be almost tongue-tied with-
out the word "capability"; we use it perhaps more often than
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
any other of our semi-technical words. Yet a little reflection
on the matter shows that we use it indiscriminately to mean
one of three quite different things: a feasible course of ac-
tion, a raw strength, and a talent or ability. Can we be sure
that we are always conveying an intended sense?*
If we do not rigorously define our terms we are likely to
find ourselves talking at cross purposes; and such discussion,
we all realize, risks being more of a fruitless dispute than an
elevated debate. This takes me to a third point.
The literature I have in mind will, among other things, be
an elevated debate. For example, I see a Major X write an
essay on the theory of indicators and print it and have it
circulated. I see a Mr. B brood over this essay and write
a review of it. I see a Commander C reading both the pre-
ceding documents and review them both. I then see a
revitalized discussion among the people of the indicator busi-
ness. I hope that they now, more than ever before, discuss
indicators within the terms of a common conceptual frame
and in a common vocabulary. From the debate in the litera-
ture and from the oral discussion I see another man coming
forward to produce an original synthesis of all that has gone
before. His summary findings will be a kind of intellectual
platform upon which the new debate can start. His platform
will be a thing of orderly and functional construction and it
will stand above the bushes and trees that once obscured the
view. It will be solid enough to have much more built upon
it and durable enough ak... I in the
oust gu Sv that no one need get bath utc
bushes and earth to examine its foundations.
Editor's Note: In our next monograph, one of Mr. Kent's colleagues, Abbot Smith,
takes up precisely this: problem in his article Capabilities in National Estimates.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
FOR OFFICIAL USE ONLY
Now if all this sounds ponderous and a drain on time, I can
only suggest that, so far, we of the Western tradition have
found no faster or more economical way of advancing our
understanding. This is the way by which the Western world
has achieved the knowledge of nature and humanity that we
now possess.
These are only three things that I would expect from this
literature. There are many others. It could and should
record such things as new techniques and methods, the history
of significant intelligence problems and accomplishments, the
nature of intelligence services of other countries, and so on.
But the three items that I have singled out remain the most
important.
There are perils of going forward in our profession without
laying down such a literature. First, there are the obvious
perils of denying our calling the advantages I have discussed
above. There is, however, another peril and one we should
heed for strictly utilitarian reasons. As things now stand,
we of the intelligence profession possess practically no per-
manent institutional memory. Our principal fund of knowl
edge rests pretty largely in our heads; other funds of knowl-
edge are scattered in bits through cubic miles of files. Whatj
happens to our profession if we are demobilized as we were
after the two world wars? What happens to it if our heads}
and files find themselves in the middle of a nuclear explosion?
The answer, I fear, is that a new beginning will have to bel
made virtually from scratch. Most of what we know will go
when we go; only a very small part will be left behind. A
ence is a reasonable insurance policy
f intelli
t
li
g
ure o
tera
against repetition of two demobilizations of intelligence that
have occurred within our memory.
FOR OFFICIAL USE ONLY
FOR OFFICIAL USE ONLY 9
In highlighting the desirability of producing a literature of
intelligence and stressing the perils of not producing one, I
do not wish to seem to close my eyes to problems and difficul-
ties.
The first of these is probably the matter of security. One
can expect the question: "Do you want to put all the secrets
of the profession in writing and bind them up in one great
book so that your enemy's success with a single target will
at once put him abreast of you?" The answer comes in two
parts. In the first place, many of the most important con-
tributions to this literature need not be classified at all. They
could be run in the,daily press and our enemies would get
no more good from them than from the usual run of articles
published in our professional journals. Surely the enemy
would benefit in some degree; he would benefit as he pres-
ently does from his reading of The Infantry Journal or Foreign
Affairs. On the other hand, another type of contribution
would deal with delicate trade secrets and would have to be
classified. But is this reason not to write or circulate it?
Every day we have to decide on the correct security pro-
cedure with respect to sensitive materials. Why should the
literature at instance be necessarily more delicate or sensi-
tive than the last cable from Paris, and why should its proper
handling be more difficult or dangerous? In this case, as
in the more familiar one of the sensitive report, we must again
equate the value of exposing many minds to a problem with
Ihn increasing danger of disclosure. The plain fact is that
.t?curity" and the advance of knowledge are in fundamental
conflict. The only reason we get anywhere is because we
do not demand either perfect security or unlimited debate
about secrets of state. We do get somewhere because the
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
10 FOR OFFICIAL USE ONLY
necessity for compromise at both ends is well and fully under-
stood.
There is another difficulty and a very practical one. How
is such a literature to be written if most or all of the potential
authors are practicing members of the profession, already
burdened with seemingly higher priority tasks? I know of
no magic formula by which a man can do two things at once.
The question that we face is the familiar one of priorities.
Surely one of the guiding principles to a solution is the de-
sirability of investing for the future. Taking Mr. X off the
current task and giving him the time to sort out his thoughts
and commit them to paper will more than repay the sacrifice
if what Mr. X puts down turns out to be an original and per-
manent contribution. If it buttons up a controversial matter
and precludes thousands of hours of subsequent discussion,
the cause has been well served. It has been well served
even though one of Mr. X's would-be consumers had to get
along without his advice on another matter. What we are
faced with in this case is nothing more complicated than the
value and pain of capital formation.
A third problem. How may the Mr. X's be paid for work-
time spent in the creation of this literature? If what has gone,
before is the fact and the Mr. X's of the calling are really
creating intelligence capital, then it seems to me that they
are entitled to their wage exactly as if engaged upon thei
regular assignments. Indeed, in logic, if what Mr. X pr
duces contributes to the solution of the next hundred prob
lems, he should be paid more than if he spent his time merely
solving the single assigned problem before him.
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
FOR OFFICIAL USE ONLY 11
Beyond these rather fundamental matters, there are hun-
dreds of other problems. If a large proportion of the Mr.
X's are sure to come from intelligence staffs, where do they
work? Are they to have secretarial help? Will they keep
regular hours? Must they be in residence? How will their
findings be reproduced? How circulated? What editorial
controls will be exercised over their output? These are really
easy questions. The hard ones are to find the Mr. X's in
the first place, and to induce them to undertake the most
difficult job of all: original creative writing.
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
12 FOR OFFICIAL USE ONLY
THE CURRENT PROGRAM FOR
AN INTELLIGENCE LITERATURE
by The Editors
WE agree with the basic ideas set forth by Mr. Kent.
We agree that there is a need for a written litera-
ture of intelligence theory and methods; that this
literature should attempt to define, criticize, and improve on
the "first principles" of, intelligence; and that this literature
can only be written by experienced officers, presenting their
own personal views. This monograph series, Studies in In-
telligence, is a first modest attempt to meet these needs.
We will, from time to time, publish articles that seem to us
to carry forward the purposes that Mr. Kent has attributed
to a professional literature. Some articles will deal with
methods of analysis and operations, some with critical defini-
tions, some with problems of organization, and some with the
special contribution of particular disciplines to the intelligence;
effort. Taken all together, we hope, these studies may get'
us started on the systematic examination of basic intelligence
theory and methods.
It is hard to pin down precisely what we mean by basi
theory, but we can identify its services to the intelligence
effort and the way it gets formulated. By theory we mea
that body of hypotheses that guides the intelligence office
in his day-to-day practical activities, that lends some co
sistency to these activities and gives him a basis for measuring
how he is doing on his job. Mr. Kent calls this body of by
FOR OFFICIAL USE ONLY
MORI/HRP THIS PAGE
FOR OFFICIAL USE ONLY 13
potheses "first principles" and says that with them as a basis,
the intelligence community makes best use of its experience
and develops professional expertness. Such first principles
are not rigid; they are always subject to change or, at least,
refinement; and they are built up - or have been, by the
academic disciplines - through a process of cumulative in-
dividual contributions by members of the profession. It is
just this building process that we want to stimulate.
What, then, is the practical upshot of this? First, the Office
of Training will act as sponsor only. Our job will be that of
generating interest in the program, getting studies written,
exploiting some studies that already exist in personal and
office files, exercising and coordinating editorial judgment,
and finally providing the publication medium for contribu-
tions to this literature. Second, the studies will in every case
be the contributions of identified Agency officers (sometimes
we will have to use pseudonyms, for security reasons, but the
general rule will remain the same) and will represent only
their own best views. And third, these views will in no case
be put forward as Agency or Office of Training doctrine.
Naturally, we will be responsible for the good sense and
factual accuracy of what we publish, but not for the sub-
stance of the arguments and criticisms and opinions ex-
pressed. We will operate on the premise that the enterprise
is worth doing but that its quality will depend entirely on the
interest of Agency people - and on their personal contribu-
tions.
Background and Charter
Over the past seven or eight years, there have been any
number of suggestions for "professionalizing" the intelligence
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
business - including everything from a trade journal to a
graduate school of intelligence. Common to all of them,
though, has been some such major premise as the one that
Mr. Kent so persuasively states above, that intelligence will
.come of age as a profession only if it recaptures its experi-
ence and the refinements of its methods in a permanent
literature. In September 1954, the Director of Training con-
vened a group of senior Agency officers to consider how best
to go about the job. This monograph series is a direct re-
sult of last year's discussion.
There are, the conferees agreed, two quite different sorts of
intelligence literature. One comprises overt material which,
whatever its stated purpose, in effect contributes to our think-
ing about intelligence and its methodology. Some of this
material is, in fact, avowedly about intelligence - in the bulk
of cases about clandestine operations. But there is still more
overt literature which can sometimes be studied with profit
by intelligence officers - books, for example, about social
science methodology or about national policy-making proc
esses. Most issues of our Studies in Intelligence, therefore
will have a bibliographic section, devoted to spotting and r
viewing some of this material. Occasionally an entire stud
will be given over to a collective discussion and critique c
a whole bloc of overt material - as, for example, a stud
now being written for us on the current state of social scienc
methodology, with the emphasis on science, and its possibli
relevance to the intelligence process.
The second major category of intelligence literature is whd
we are primarily concerned with. This is the material the
can only be prepared by experienced intelligence office
and will usually be classified because of references to tl%
MORI/HRP THIS PAGE
FOR OFFICIAL USE ONLY 15
mission and product of the intelligence community. Many
officers have formulated, even though vaguely, certain prin-
ciples or methods drawn from their experience in dealing
with a succession of problems or cases. Or, conversely, they
have come to feel that their experience challenges the valid-
ity of a commonly-held concept. These ideas occasionally
become the subject of an office memorandum or get dis-
cussed informally; more often they simply evaporate in the
heat of current business. In any event, they are not care-
fully thought through and then presented for the enlighten-
ment and serious consideration of interested associates. We
hope that the Studies series will provide a vehicle through
which such experienced officers can systematically speak
their minds.
Organization and Procedure
Our current procedure for obtaining contributions to the
series is simply to encourage various Agency officers to pre-
pare studies about the problems in which they are especially
expert, which they deal with continuously in their work, and
which they think are so fundamental to so much of the busi-
ness of intelligence as to be appropriate subjects for a basic
literature. Thus, when we wanted an article on "capabili-
ties" analysis in the estimative process, we turned to Abbot
Smith of the Office of National Estimates - whose study, by
the way, will appear soon in this series.
When a study arrives at our desk, we first send it out to a
number of other men whose experience qualifies them'as
critics and advisers, and ask: Is this piece worth publishing?
Should it be revised? If so, precisely how? After the advice
and criticism is in, however, it is up to the author to decide
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
what he wants to sign his name to; and it is up to us to decide
whether to publish at all.
When we send out articles for comment, we are certainly
not seeking "coordination." We will be happy enough if
our preliminary readers will agree that the author's point
of view is sound and knowledgeable; we surely do not count
on agreement about substance. Nor do we imagine that one
article will necessarily exhaust a subject. When we can pre-
dict that a subject will clearly break down into two or more
conflicting points of view, we will try to find representatives
of each and publish a symposium. Usually, though, we will
depend on reader-reactions, in the form of letters to the edi-
tor (which will be published) and suggestions for further
studies. To borrow again from Mr. Kent's terminology: we
will never avoid debate but concentrate on keeping it "eleva-
ted."
Our dependence on soliciting studies is, we hope, only tem-
porary. If the project is worth anything, one proof will pre-
sumably be the amount of interest it arouses - and the num-
ber of unsolicited suggestions received. The address is 2204
Alcott Hall; or call C. M. Lichenstein at ext. 2428. We wel
come comments on the first studies,and prospectuses for more.
To be more specific about our publication program, thes
are some of the studies now underway, all of which shout
appear before the year is out:
"Capabilities" in National Intelligence Estimates
The Nature and Role of Economic Intelligence - an
Some of its Methods of Analysis
MORI/HRP THIS PAGE
The Scientific Method and the Intelligence Process
Administration in Intelligence
Theory of Indicators - and a Case Study
"Readability" in Intelligence Publications
Scanning this list, one will see that we are not in the sub-
stantive intelligence business; we are not competing with any
producing Agency office; indeed, we will not publish finished
intelligence at all.
On the contrary, we want to publish studies that could not
possibly appear as finished intelligence analysis or as opera-
tional support under official Agency auspices (but might be
prepared as preliminary, methodological working papers);
studies that deal with the way an intelligence officer does his
job, with the techniques and methods he uses. To each pro-
posal we will apply the criterion: will this paper contribute to
the professional theoretic literature, as best it can be defined?
Rather than prepare a list of possible study topics, then, we
want to review each prospectus or manuscript that reaches
us against this criterion.
After we have been publishing for a time, we hope that our
own experience and the criticism of many readers will have
sharpened understanding of the exact nature of this basic
literature and how best to get it written. Certainly the sub-
stance of the publications and our own operating procedures
are equally subject to change for the better - and, in both
cases, the major share of the burden seems to us to be on
the people who have made intelligence a respected profes-
sion.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
MORI/CDF THIS PAGE
MORI/CDF THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
EDITORS INTRODUCTION
ecting the;Ivataonal
material contains, information aff
Defense of the',United States" within the';meaning o` 2lie
,espionage laws, :Title 18, USC,7Sees. 793 and 794, the trans-
'mission or`revelation of which to an unauthorized person~is
prohibited by law.
Office ??of. Training-
Subsequent-issues willbe disseminated end y
s. pg P.
~ he pg ~y~~'~ol~make~ sure of frecei~ o
ecure ,extra copies, please call; the extensio
MORI/CDF THIS PAGE
sag=
Central~Intelligence Agency ~i of the Otfice~of sTraininng
N September, the Office of Training issued the introductory
number of Studies in Intelligence. Our purpose, we said,
was to stimulate thinking and writing about the f unda-
mentals of intelligence work, and to sponsor the beginnings of
a professional intelligence literature. We especially empha-
sized two requirements basic to the production of such a litera-
ture: first, all that we publish will be entirely unofficial and
will represent only the opinions of the individual author,
second, the success of the project will depend on participation
by the whole intelligence community. Successive Studies, that
is, to sav, will appear only as worthwhile manuscripts reach our
ii ly
d a a what esk; and we will be able to judge himpact of w`11aL we pub-
iish only as we receive reader comments.
presenting this issue on "capabilities" we call your atte n
tion to a concept whose applications extend to nearly every
aspect of intelligence work. Just about everyone, at one time
or another, is in the capabilities business, from the case officer
who keeps current and reports on the "capabilities" of a nation-
al Communist Party to the Board of National Estimates which
turns out exhaustive studies on the "capabilities" of the Soviet
Bloc. One of the classic definitions of intelligence is, indeed,
"the analysis of the capabilities and vulnerabilities of foreign
countries, relevant to US security interests." Both authors
tackle the subject at its most basic: what do we mean by the
word and, if our meaning is not always clear and consistent,
what should we mean? What experience do we draw on in
analyzing "capabilities" and how, in specific cases, does the
R llai'cic r,rnnnoA9
born Abbot E. Smith and Harold D. Kehm bring to bear on
the subject an abundance of experience in intelligence (spe-
ally r. capabilities analysis) and related fields Mr. Smith,
it Rhodes Scholar and a distinguished historian, has taught at
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Bard College and Columbia University. During World War
II, and immediately after, he served in the US Navy in a variety
of roles: as Acting Chief, Naval Division, Allied Command in
Austria; as Chief of the Historical Section, US Naval Forces
in Europe; and as a member of the Historical Section, Joint
Chiefs of Staff. Since 1943, Mr. Smith has been with CIA.
Col. Kehm has been an Instructor in Economics and Govern-
ment at West Point and Chief of Instructor Training at the
Command and General Staff College, G-2 of the Ninth Army
during World War II, Army Attache in Dublin, and Assistant
Commandant (i.e., Chief of the School) at the Strategic In-
telligence School. Col. Kehm joined this Agency in 1954.
The two articles that follow by no means exhaust the subject
at hand - it is much too broad and involves too many side-
issues for that. At least two directly related problems, each
worth a Study in itself, have occurred to us as we have reflected
on Mr. Smith's and Col. Kehm's contributions. One is the
problem of the special characteristics of national, as distinct
from departmental, intelligence. To put it in the form of a
question: to what extent is the experience and the methodology
of, e.g., military intelligence directly applicable to the produc-
tion of national intelligence? The terminology has carried
over, to be sure; but in Mr. Smith's and Col. Kehm's articles
there are differences in usage of the capabilities concept that
may result in part from basic differences in the problems the
national and the military intelligence officer are asked to solve.
Then, too, there is the problem, raised in both articles, of the
lack of a national G-3 - which may, again, complicate the
process of applying the systematic and time-tested methodology
of the military intelligence officer to national intelligence. And
surely there are many other problems of "capabilities" that
could usefully be addressed in subsequent issues of this series;
these are but two of the more obvious.
We invite suggestions and prospectuses, therefore, for some
of these unwritten Studies and comments on the present one -
comments which we would like also to publish in subsequent 4
SECRET
issues. To repeat something we think hears a lot of repeat-
ing: if indeed these Studies in Intelligence are to help in the
airing of intelligence principles and methods, in the recaptur-
ing of experience, and eventually in the building of authorita-
tive doctrine, then we are going to need the advice and the
participation of every member of the intelligence profession to
do the job well.
CONF1DENT'P
Spow
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
NOTES ON "CAPABILITIES" IN
NATIONAL INTELLIGENCE
by Abbot E. Smith
W HEN CIA was established with the mission of producing
"national" intelligence it perforce drew heavily for
doctrine upon the military intelligence agencies. Over
the years, the intelligence organizations of the armed forces
had developed a well-tested routine. Formulas were available
to meet various requirements. Agreement had gradually been
reached on what needed to be known about the enemy, what
data were necessary for the estimate, why they were necessary,
and how they could most usefully be presented. CIA had no
counterpart to this doctrine. It therefore frequently borrowed
from the military, and in no instance was this borrowing more
conspicuous than in the matter of "capabilities."
The doctrine of enemy capabilities is one of the most charac-
teristic and useful that military intelligence has to offer. A
capability is a course of action or a faculty for development
which lies within the capacity of the person or thing concerned.
More particularly, in military intelligence, enemy capabilities
are courses of action of which the enemy is physically capable
and which would, if adopted and carried through, affect our
own commander's mission.* In short, a list of enemy capabili-
"'capabilities, enemy -Those courses of action of which the enemy
is physically capable and which if adopted will affect the accom-
plishment of our mission. The term "capabilities" includes not
only the genera! courses of action open to the enemy such as attack,
defense, or withdrawal but also all the particular courses of action
possible under each general course of action. "Enemy capabilities"
are considered in the light of all known factors affecting military
operations including time, space, weather, terrain, and the strength
and disposition of enemy forces . . ." Dictionary of United States
Military Terms for Joint Usage, issued by the Joint Chiefs of Staff.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
2 SECRET
ties is a list of the things that the enemy can do. It is there-
fore apt to be the most significant part of a military intelligence
officer's "Estimate of the Enemy Situation."
It is true, of course, that a military intelligence officer col-
lects and transmits to his commander a great deal of other
information. He reports on the weather, terrain, and com-
munications in the zone of operations. He may set forth the
politics and economics of the area. He collects and evaluates
data on the enemy's order of battle, logistical apparatus, equip-
ment, weapons, morale, training and the like. All this is made
known to the commander, but it is still not a statement of
enemy capabilities. Only when the intelligence officer has ac-
quired all this information, and constructively brooded over it,
can he set about describing the courses of action open to the
enemy. It is this list of capabilities that tells the commander
what, under the conditions existing in the area, the enemy can
do with his troops, his weapons, and his equipment to affect the
commander's own mission. The enumeration and description
of enemy capabilities is the ultimate, or at least the penulti-
mate, goal of military intelligence. It is one of the character-
istic modes to which the great mass of intelligence information
available is bent, in order to give the commander the knowledge
of the enemy he needs to plan his own operations.
Adaptation of this doctrine to the requirements of national
intelligence presents at first no real difficulty. Courses of ac-
tion may be attributed to persons, organizations, parties, na-
tions, or groups of nations as well as to military units, and to
friendly or neutral, as well as to enemy, powers. They may be
political, economic, psychological, diplomatic, and so on, as well
as military. It is true that a national intelligence estimate*
is not made for a military commander with a clearly defined
* Throughout this paper the term "national intelligence estimate"
is used generally to mean not just the solemnly coordinated "Na-
tional Intelligence Estimates" approved by the Intelligence Advisory
Committee, but any estimate, great or small, made by any office or
person producing national intelligence.
MORI/HRP THIS PAGE
mission, to which enemy capabilities may be referred to ascer-
tain if they do in fact "affect" the carrying-out of that mission.
An equivalent for the commander's mission is not far to seek,
however, since national intelligence is obviously concerned only
with foreign courses of action which may affect the policies or
interests - above all the security interests - of the United
States. It is by no means as easy to be clear about all the poli-
cies and interests of the United States, and to perceive what
might affect them, as it is to understand the mission of a mili-
tary commander, which is supposed to be unequivocally stated
in a directive from higher authority. But this is one of the
reasons why a national intelligence estimate is apt to be more
difficult to prepare than a military estimate of an enemy situa-
tion.
In national intelligence, then, capabilities may be defined as
courses of action within the power of a foreign nation or organ-
ization which would, if carried out, affect the security interests
of the United States.
It is probably unnecessary to argue that statements of capa-
bilities are useful as a means of organizing and presenting na-
tional intelligence. The parallel with military intelligence doc-
trine seems perfectly sound. High policy-makers doubtless
want to be supplied with authoritative descriptions and anal-
yses of the politics, economics, and military establishments of
various foreign nations, together with explanations of the ob-
jectives, policies, and habitual modes of action of these nations.
They need to have the best possible statistics, diagrams, pic-
tures, and data in general. But when all the labor and re-
search has been finished, the results collated and criticized,
and the conclusions written down, it will still be worthwhile to
go on to a statement of what each foreign nation or organiza-
tion can do to affect the interests of the United States. This
is the statement of capabilities.
In recognizing, formulating, testing, and presenting foreign
capabilities, intelligence doctrine comes into its own. Apart
from the special function of intelligence operations in collecting
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
4 SECRET
data, most of the preliminary spade-work for intelligence esti-
mates is the province of other disciplines than that specifically
of intelligence. This spade-work of course takes nine-tenths of
the time, trouble, and space devoted to any estimate. Political
scientists analyze the structure of government and politics in a
foreign state; economists lay bare its economic situation; order-
of-battle men reveal the condition of the military establish-
ment; sociologists, historians, philosophers, natural scientists,
and all manner of experts make their contribution. When all
this has been done it is the peculiar function of intelligence
itself to see that the learning and wisdom of experts is directed
towards determining what the foreign nation can do to affect
US interests. Thereby the major disciplines of social and
natural science are turned to the special requirements of
intelligence estimates.
Let us be careful not to confuse this with the function of
prophecy. To predict what a foreign nation will do is a neces-
sary and useful pursuit, albeit dangerous; it rests on knowledge,
judgment, experience, divination, and luck. To set forth what
a nation can do is a different matter. One still needs judgment,
experience and luck as well as knowledge, but soothsaying is
reduced to a minimum. There is an element of the scientific.
The job can be taught, and its techniques refined. It can be
reduced to doctrine.
Generally speaking, in military usage an enemy capability is .
stated without reference to the possible counteractions which
one's own commander may devise to offset or prevent such
action. The Navy's handbook entitled Sound Military Decision
puts it this way (italics added) : "Capabilities . . . indicate
actions which the force concerned, unless forestalled or pre-
vented from taking such actions, has the capacity to carry Out."
MORI/HRP THIS PAGE
Here are three examples:
a. The Bloc has the capability to launch large-scale, short-
haul amphibious operations in the Baltic and Black Seas.
b. The USSR has the capability to launch general war.
c. The Chinese Communists have the capability to commit
and to support approximately 150,000 troops in Indochina.
These statements give no estimate of what the effects or
results of any of these courses of action might be. There is
no indication for example that the United States or some other
power might be able to make it difficult or impossible for the
Chinese Communists to support 150,000 troops in Indochina,
or that the West might possess such strength that a Soviet
decision to launch general war would be tantamount to suicide.
The statements simply lay down what the nations concerned
could do, without regard to any possible opposition or counter-
action. Such unopposed capabilities are frequently referred
to as "gross" or "raw" capabilities. They are the kind of
enemy capabilities which are reported to a military commander
by his G-2, in the "Estimate of the Enemy Situation."
The high policy-makers for whom national intelligence is
designed, however, are not in the comparatively simple position
of military commanders facing an enemy. They have broader
fields to cover, and more numerous problems to face. They
need to have a picture of the security situation in the world
as a whole and in various areas of the world. This picture
ought to show not only the multifarious forces which exist, but
also the probable resultants of these forces as they act upon
each other, or as they might act upon each other if they were
set in motion. The policy-makers need, in short, to know about
net capabilities, not merely about gross or raw capabilities.
This is well understood and accepted as long as the courses
of action of foreign nations alone are concerned. Nobody
would think of enumerating the capabilities of France, for
example, without giving due consideration to the frequently
opposing capabilities of Germany, and to the tangential capa-
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
bilities of Great Britain and other powers. Even in the purely
military sphere, statements of net capabilities occur in national
estimates. For example:
a. In Israel, an army of 49,000... is capable of defeating any
of its immediate neighbors.
b. The Chinese Communists have the capability for conquer-
ing Burma.
c. We believe that the Chinese Communists are capable of
taking the island of Quemoy if opposed by Chinese Nation-
alist forces only.
It is an intricate and difficult operation even to attempt to
work out the probable resultants of the enormous forces actu-
ally or potentially at work in the world - political, economic,
military, and the like. Without such an operation, however -
sometimes called "war-gaming" when limited strictly to the
military sphere - national intelligence estimates of capabilities
would lose much of their usefulness for the particular purpose
they are designed to serve.
Obviously no estimate of the security situation anywhere
in the world will be worth much unless the capabilities of the
United States are taken into account and their effect weighed.
At this point, however, grave practical difficulties arise. We
of the intelligence community are solemnly warned that we
must not "G-2 our own policy." Military authorities are
shocked at the suggestion that we should indulge in "war-
gaming." We are told that it is the function of the commander,
not of the intelligence officer, to decide what counteraction to
adopt against enemy capabilities, and to judge what the success
of such counteraction may be. It is pointed out that no ade-
quate estimate of net military capabilities can be made without
a full knowledge of US war plans, and a long and highly tech-
nical exercise in war-gaming by large numbers of qualified
experts. Since intelligence agencies as such quite properly
have no knowledge of US war plans, and possess no elaborate
machinery for war-gaming, they are estopped from making
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
SECRET 7
an estimate of net capabilities where US forces are significantly
involved. As a result there is, for instance, no statement in
any national intelligence estimate of how the military security
situation on the continent of Europe really stands, i. e., of the
probable net capabilities of Soviet forces against the opposition
they would be likely to meet if they attempted an invasion of
the continent.
This state of affairs is unfortunate, and the value of national
intelligence estimates is thereby reduced below what it ought
to be. The difficulty is really not one of intelligence doctrine,
however. Practically nobody doubts that high policy-makers
ought to be supplied with estimates of net capabilities even in
situations where the US is actively engaged. It is agreed that
they ought to have the best possible opinion on the security
situation on the continent of Europe, and that they must be
informed not merely of the gross capabilities of the USSR
to launch air and other attacks on the US (the subject of an
annual National Intelligence Estimate) but of what the USSR
could probably accomplish by such an attack against the
defenses that the US and its allies would put up. In one way
or another policy-makers get such estimates of net capabilities,
even if they have sometimes to make them themselves, off
the cuff.
The question is, then, not whether estimates of net capa-
bilities are legitimate requirements, but simply who shall make
them. This problem is outside the scope of a paper on intelli-
gence doctrine. It may be suggested, however, that the diffi-
culty has probably been somewhat exaggerated. The jealous
prohibition of "war-gaming," on grounds that to conduct it
requires a knowledge of US war plans and an enormous appa-
ratus with numerous -personnel, is overdone. In four out of five
situations where an estimate of net military capabilities is
needed the judgment of wise and experienced military men,
based on only a general knowledge of US war plans, is likely
to be about as useful as the most elaborate and protracted
piece of war-gaming. Such exercises have too often given the
MORI/HRP THIS PAGE
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
wrong answer - they are really no more dependable as guides
to the outcome of future wars than research in economics is
dependable as an indicator of the future behavior of the stock
market. This does not mean, of course, that economics and
war-gaming are useless pursuits.
Gradually, indeed, the difficulties respecting estimates of net
capabilities are disappearing. In the most critical situations -
air attack on the United States, for example, and perhaps the
security situation in Europe - it may be necessary to establish
special machinery for the most careful playing-out of the
problems and ascertainment of net capabilities. In less critical
situations the trouble is solving itself. Military men are be-
coming a little less shy of making an educated guess as to net
capabilities, even when US forces are involved, and the com-
munity is not as distressed as it used to be at the accusation of
"G-2-ing US policy." A doctrine is gradually being evolved by
trial and error, which is as it should be. Some day it may be
desirable to commit the evolved doctrine to writing, but the
time has not yet arrived.
Of course any foreign nation of consequence is physically
capable of a vast number of courses of action which would
affect the security interests of the United States. One task
of intelligence (after the spade-work is complete) is to recog-
nize these capabilities; another is to test them against known
facts to make certain that they are real and not imaginary;
a third is to test them one against another to see how many
could be carried out simultaneously, and how many may be
mutually exclusive; a fourth is to work out in reasonable detail
the implications, for the nation concerned and for the United
States, of the actual implementation of each important capa-
bility. I propose to pass over all these tasks without further
discussion, and to concentrate on the problem of selecting from
SECRET 9
among the capabilities those which are to be included in the
formal estimate. For even after all the testing is finished
there will still remain far too many capabilities to put into
any document of reasonable size. Considerations of space,
time, and the patience of readers make it imperative that some
principles of exclusion be adopted, so that the list of capabili-
ties presented will be useful rather than merely exhaustive.
Capabilities are excluded from national estimates for one of
two reasons: either because they are judged unlikely to be
actually adopted and carried through, or because they are con-
sidered to be so insignificant that they could be implemented
without more than minor effect on the security interests of the
United States. For short we may say that they are excluded
on grounds either of improbability or of unimportance.
The second of these criteria does not require much discus-
sion. Clearly it would be a waste of time and paper to fill a
national estimate with lists of courses of action which, even if
carried out, would affect the security interests of the United
States only to an insignificant degree. One applies common
sense in this matter, and forthwith rejects a great number of
capabilities from further consideration. Along with common
sense, however, there ought always to be plenty of specialized
knowledge available. Everyone knows that an expert can
sometimes point out major significance in things which are
to the uninformed view negligible, and conversely that experts
will sometimes inflate the importance of things which common
sense and general knowledge can see in juster proportion. Out
of discussion and argument on these matters comes the best
verdict as to the importance or unimportance of a given
foreign capability, and the best guidance as to whether it
should be put into the formal estimate.
To reject any foreign capability because we judge it unlikely
to be implemented is a more serious and difficult matter. Here
indeed we part company with military doctrine, which frowns
upon the exclusion from an estimate of any enemy capabilities
whatever, and especially condemns any exclusion on grounds
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
10 SECRET
of improbability. There has been much debate, among the
military, on whether an intelligence officer should presume to
put into his formal estimate an opinion as to which of the
enemy capabilities listed is most likely to be implemented. It
has been said that such a judgment is for the commander
alone to make, and some have even held that the commander
himself must not make it, but must treat all enemy capabilities
as if they were sure to be carried through, and must prepare
to deal with them all. This latter doctrine is somewhat aca-
demic. It is doubtful that any intelligence officer, or any
commander worth his salt, has ever acted strictly in accordance
with it. Yet it remains that according to the more rigorous
teachings of military intelligence no enemy capability of any
consequence may be omitted from the list presented to the
commander. The disasters which can result from even a care-
fully considered exclusion have been frequently pointed out.
Nevertheless, in a national intelligence estimate we must for
the reasons already stated exclude many foreign capabilities
because we judge them unlikely to be carried out. The unlike-
lihood is in turn generally established on one or more of three
grounds, namely, that implementation of the capability (a)
old objec-
would be unrelated to, or incompatible with, national
tives tives of the country under consideration; (b)
occompulsions
counter to the political, moral, or psychological
~ou d
under which the nation, or its rulers, operate;
entail consequences so adverse as to be unprofitable.
The most obvious capabilities to exclude are those which, if
implemented, would serve no objective of the nation under
consideration, or would clearly run counter to some of that
nation's objectives. Thus we do not bother about the possi-
bility that the British might conquer Iceland, although they
certainly could do so and if they did US security interests
would be affected. The conquest of Iceland, however, would
serve no British objective that we know of, at least in time of
peace. Again, it is clearly within the power of the USSR to
give up its Satellites, renounce its connections with Commu-
MORI/HRP THIS PAGE
SECRET 11
nist China, and retire modestly into isolation. Or the British
might, in order to improve their economic condition, abandon
all armaments and cease to be a world power. We do not give
such capabilities serious consideration, however, because we
believe them manifestly contrary to the fundamental aims of
the Soviets and British respectively. By applying this sort of
standard we can immediately reject a great number of courses
of action which lie within the power of the nation concerned
and which would affect US security interests.
One must be careful in using this test, however, for national
objectives change, sometimes with changes in government,
sometimes without. It is, for example, impossible to be sure
about the objectives which will determine West German policy
in years to come. Even the Soviets do not always appear to
the Western view to act in such a fashion as to serve what we
estimate to be their real aims. Moreover, all nations have
various objectives, many of which are to some degree incom-
patible with each other. Sometimes one is governing, some-
times another. Nations can even pursue simultaneously several
conflicting objectives, to the confusion of their own citizenry
,LS well as of foreign intelligence officers. We must be very
certain, before rejecting a foreign capability as incompatible
with a national objective, that the objective is genuine, deeply-
felt, and virtually certain to govern the nation's courses of
action.
The political, moral, or psychological compulsions which
operate on a nation, or on its rulers, make the implementation
f some of that nation's physical capabilities unlikely or even
impossible. Thus, for example, it would probably be judged
that the US is unlikely to undertake a strictly "preventive"
war against the USSR because such an action, under any
foreseeable US government, would be politically and morally
unthinkable. It may similarly be true that the Soviet rulers
are psychologically unable to establish a genuine state of
peaceful coexistence with capitalist states even though they
mac proclaim their desire to do so and may judge such a
MORI/HRP THIS
PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
12 SECRET
course of action conducive to the ultimate aims of Communism.
There are some things that nations cannot do, despite the
fact that they are physically capable of doing them and might
serve their national objectives thereby.
To be sure, if a nation is politically, morally, or psychologi-
cally incapable of pursuing a given course of action that
course of action is not a capability at all, and we need not
worry about it. The trouble is, however, that while physical
incapabilities can generally be pretty satisfactorily established
the same is rarely true of political, moral, or psychological
incapabilities. One must depend more on judgment and less
upon demonstrable certainty for an estimate in the matter.
Not many would have estimated, before the fact, that Tito
would be psychologically capable of turning against Stalin,
or that the Germans would be morally capable of supporting
Hitler, or that the United States would be politically capable
of abandoning isolationism. Experience warns us against un-
due confidence in our estimates of national character, and it
will be safer to consider as capabilities all courses of action
which a nation is physically able to carry through, rejecting
many as improbable but none as impossible.
Finally, we reject from our estimate those capabilities which
would, if implemented, lead to such adverse consequences as
to be unprofitable. There are, curiously enough, very few
foreign capabilities which will pass the tests already mentioned,!
and then have to be excluded on this ground. This is because
most courses of action having indubitably dire consequences
will by reason of that fact alone run counter to the objectives
or to the political, moral, or psychological compulsions of the
nation. Those few which are left are generally military in
nature and are apt to be so important that we include them
in the estimate anyway. Thus it is clear that general w
with the US would be hazardous and perhaps disastrous fo:
the USSR. It therefore seems highly improbable that th
Soviets will deliberately run grave risks of involving themselves
in such a war, yet no national estimate on the USSR would'
MORI/HRP THIS PAGE
SECRET 13
omit mention of the capabilities of that nation for conducting
war with the US. The same holds true for the capabilities of
the Nationalist Chinese to invade the mainland, or of the
South Koreans to attack North Korea. We may judge such
capabilities improbable of implementation, but we do not
exclude them from our estimate.
By applying the tests of importance and of probability, as
described above, the vast number of capabilities of any foreign
nation will speedily be reduced to manageable proportions.
The process of exclusion will at first be almost unconscious -
most capabilities will be rejected forthwith, without doubt or
debate. When this stage has been accomplished, however,
there will still remain a formidably long list which will require
more serious consideration. Exclusion becomes more difficult,
and begins to require longer discussion and maturer judgment.
The same criteria of choice continue valid, but are applied with
more deliberation. This is the point at which preparation of
the estimate gets interesting, for the choice of capabilities to
include or exclude may prove to be the most crucial decision
made during the estimating process.
Though we have departed from the military doctrine in
allowing a rejection of capabilities judged unlikely of imple-
mentation, we may still return to it for an important lesson.
Like the military commander, the high policy-maker is entitled
to something more than intelligence's opinion of what foreign
nations will probably do. He is entitled to be informed of
various reasonable alternative possibilities, and to be given
some discussion of these alternatives - of their apparent ad-
vantages and disadvantages, and of the reasons why intelli-
gence deems them respectively to be less or more likely of
implementation. National estimates sometimes discuss only
the particular foreign capabilities which the intelligence com-
munity in its wisdom believes will actually be carried through.
'I'bis is going too far in exclusion. Intelligence must winnow
the mass of capabilities down to two or three or half a dozen
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
14 SECRET
in each situation examined, but it is the responsibility of policy-
makers, not of intelligence agencies, to decide which among
these few last alternatives shall in fact constitute the intelli-
gence basis for US policy.
Looking back over old national estimates one is apt to feel
that the borrowing of military terminology was sometimes a
little over-enthusiastic. The word "capability," for example,
offers an almost irresistible temptation to all of us who compose
governmental gobbledegook. It is a long, abstract noun, of
Latin derivation, and it has a pleasing air of technicality and
precision. It will appear to lend portentousness to an other-
wise simple statement. Perhaps this is why the word appears
in estimates so frequently, unnecessarily, and sometimes even
incorrectly.
One trouble is that the word has a perfectly good, non-
technical meaning, signifying a quality, capacity, or faculty
capable of development. It is commoner in the plural, when
it usually denotes in a general way the potentialities of the
possessor, as when we say that a man "has good capabilities."
This usage is frequent in estimates:
a. The air defense capabilities of the Bloc have increased
substantially since 1945.
b. Chinese Communist and North Korean capabilities in
North Korea have increased substantially.
c. The capabilities of the new fighter aircraft are superior to,
those of the old.
No valid objection can be taken to these examples. Indeed,
the usage is virtually the same as that of the technical term,:
for the statements are about the things that the possessors of.
the capabilities can do.
MORI/HRP THIS PAGE
One can find, however, a good many examples of slipshod
usage:
a. Satellite capabilities for attack on Greece and particu-
larly on Turkey are too limited for conquest of those
countries.
b. The Tudeh Party's capabilities for gaining control of Iran
by default are almost certain to increase if the oil dispute
is not settled.
There is no good reason for using the word "capabilities" in
either of these statements; in the first the word should prob-
ably be "resources," in the second, "chances" or "prospects."
If one really insists on talking about capabilities then the
statements ought to be rephrased: "The Satellites are not
capable of conquering Greece or Turkey," and "If the oil dis-
pute is not settled, conditions in Iran will be such that the
Tudeh Party may acquire the capability to gain control of the
country."
It will be perceived that the immediately foregoing examples
are statements of net capabilities, and it is in connection with
such statements that imprecise drafting most frequently
occurs. It must be remembered that in a relationship be-
tween two nations (or other organizations) the gross capa-
bilities of one side can be increased or decreased only by an
increase or decrease in the strength, resources, skills, etc., of
that side; what happens on the other side is irrelevant. The
net capabilities of one side, however, may be altered either by
a change in its own strengths and resources or by a change
In those of the other side. For example, suppose that the
strengths and resources of the United States and the USSR
both increase in the same proportion. Then the gross capa-
bilities of each side will have increased, but the net capabilities
will have remained unchanged. But, if the USSR should grow
weaker, while the United States made no change in its strength,
then the net capabilities of the United States would have in-
creased although its gross capabilities remained unchanged.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
This is simple enough, but it needs to be understood if
drafting is to be accurate and clear. Consider the following
example:
In South Korea and Taiwan where US commitments pro-
vide both physical security and political support of the
established regimes, present Communist capabilities for
political warfare are extremely small. If the US commit-
ment and physical protection were withdrawn for any
reason, substantial and early Communist political warfare
successes almost certainly would occur.
The first of the two sentences in this quotation can only be
understood as a statement concerning gross capabilities, al-
though to be sure the word is used in its non-technical sense.
But the second sentence reveals that Communist gross capa-
bilities, far from being "small," are in fact very considerable.
The two sentences together constitute a statement of net capa-
bilities, but the drafting is poor. Perhaps a rule to govern this
problem may be formulated in this way: when the word "capa-
bility" or "capabilities" is used in its non-technical sense,
signifying in a general way the qualities, faculties, or potential
of the possessor, it must be used only to refer to gross, and
never to net capabilities. If there is any question, doubt or
difficulty, the word ought to be avoided and a synonym chosen.
Finally, even when using the word in its technical meaning
of a specific course of action, the drafter ought always to make
clear whether he is referring to gross or net capability. For
example:
We estimate that the armed forces of the USSR have
a
.
the capability of overrunning continental Europe within
a relatively short period:
b. The Party almost certainly lacks the capability for
seizing control of the Japanese government during the
period of this estimate.
The first of these statements is unclear because the word
'
.
"overrunning" does not indicate beyond doubt (as "conquer
MORI/HRP THIS PAGE
SECRET 17
or "defeat" do in some examples previously quoted) whether
the statement is or is not one of net capability. Does the
sentence mean that the armies of the USSR can overrun Europe
against all the opposition that the West may put up? Or does
it mean only that the USSR has enough men and logistical
apparatus to spread into all of continental Europe within a
relatively short period if unopposed? The second example is
clearer, but still it does not indicate beyond doubt whether the
Party is unable to seize power because the Japanese govern-
ment is strong enough to prevent it, or whether the Party
simply lacks the men and talent to take over the job of govern-
ing Japan even if no one opposed its doing so.
Apart from such suggestions for clarity in drafting as those
given above, it would be premature to lay down rules for the
statement of capabilities in a national intelligence estimate.
Sometimes it may be desirable to list them seriatim, as the
military generally do in their estimates of the enemy situation.
This might be a wholesome exercise while drafting an estimate
even if it were not retained in the -final version, for it would
tend to promote precision, to reveal inter-relationships and
produce groupings of related capabilities, and thus to prevent
the indiscriminate scattering through an estimate of state-
ments of capabilities in bits and pieces. On the other hand,
the number and complexity of courses of action which have
to be presented may often be so great that extensive listing
tisouid be tedious, and attempts at grouping misleading. A
connected essay (in which, incidentally, the word capability
()r capabilities need never appear) may convey the material far
more adequately.
These matters will be improved by experimentation, and by
the talent of those who draft estimates. Improvement is worth
trying for, in this as in other aspects of estimating capabilities.
It is a great and responsible task to survey the whole political,
economic, and military strengths of a nation, to ascertain its
objectives and the moral and political compulsions that govern
its conduct, to weigh all these matters in the light of that
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
18 SECRET
nation's relation to other nations, to perceive what that nation
could do to affect the security interests of the United States,
and to select from among these manifold courses of action those
sufficiently important and feasible to be included in a national
estimate. The techniques of this task are still in a formative
stage. They will develop through experience, through trial
and error, through discussion and argument, and perhaps,
from time to time, through purely theoretical and doctrinal
investigation.
NOTES ON SOME ASPECTS OF
INTELLIGENCE ESTIMATES
by Harold D. Kehm
MEMBERS of the intelligence community will obviously
find useful reading in the articles by Abbot Smith and
Col. Kirtland.* These studies deserve the attention of
other groups as well. They are of particular value to military
commanders and planners and to their civilian counterparts
in both government and private life. The executive and the
planner are the prime consumers of the intelligence product.
Furthermore, since they and not the intelligence officer are
ultimately responsible for action taken, they are and should be
the sharpest critics of that product.
These consumers, therefore, need to understand the various
kinds of approaches which the intelligence officer can make
to his problem. In consultation with him, they should develop
an agreed approach - embodying doctrines either as discussed
in our military and other staff manuals or possibly as modified
by ideas developed in these papers.
Business executives and planners were mentioned above
along with military and government officials because study of
modern business organization and practice makes it quite clear
that the more effective enterprises engage in intelligence
activities in one form or another.
To bring out the parallel with national and military intelli-
gence, we may note that business intelligence comprises eval-
.iated information concerning such matters as: the actual
and potential users of the goods and services the business
produces; the actions and plans of competitors; related goods
and services; and other factors which bear on the production,
marketing, and use of the product. Among the "intelligence
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
activities" in which most business organizations engage we
can include market analysis, research and development, and
the collection of general business information.
Market analysis is essentially an intelligence activity, for
it covers not only what the product may or might do but also
what other firms and products may do or are doing. Credit
information on firms and individuals is perhaps the most direct
form of intelligence used by business.
Research and development is an intelligence activity in the
sense that it yields information on which to gauge the value
of one's own product as well as that of actual and potential
competitors. Research and development have become so im-
portant that investment analysts now consider the size and
quality of this effort an important factor in determining the
value of a security.
Finally, no business of any stature can plan without giving
at least a quick glance at political, economic, and sociological
data. It is inconceivable that either Ford or the UAW in 1954
planned for 1955 without considering international affairs, the
domestic political situation, and the sociological "climate"
which might make it propitious to raise the issue of the guar-
anteed annual wage. The tremendous growth in the number
of trade and commercial publications is an indication of the
interest in business intelligence information.
This is not the proper place to pursue this matter further
and discuss whether or not business would improve its lot by
openly recognizing its intelligence requirement and organizing
more specifically for it. It is useful to note, however, that
World War I taught business leaders the value of the line
and staff principle of organization and that World War II
has already given them clear object lessons in operations;;
analysis and on research and development. "Business intelli-,
gence," full-fledged, may well be the next important step.
It has seemed worthwhile to mention this point because we
want to go along with Mr. Smith who believes that military"
MORI/HRP THIS PAGE
SECRET 21
intelligence doctrine has application in national policy proc-
esses. In fact, we want to go further and assert that the basic
concepts - not necessarily all the detailed precepts and pro-
cedures - have application to any form of human activity:
political, economic, scientific, or sociological.
There is some reason to suspect that both Mr. Smith and
Colonel Kirtland have misinterpreted or misunderstood some
of these basic concepts. We propose to deal with these mis-
understandings as they come up in our discussion of the two
papers. At this point, it is useful to cover one matter which
both seem to have failed to keep clearly in mind. It is the
fact that both the intelligence officer and the commander (or
policy-maker) are in the estimating business.
The Intelligence Function and the Command Function
The intelligence officer is the "expert" on the enemy. Ac-
cordingly, he is charged with giving the commander, the staff,
and subordinate commands the best information and esti-
mates on the enemy situation. The end product of his estimate
is enemy capabilities and - let us not forget - where available
information provides a basis for such judgment, the relative
probability of adoption of them.*
This is a full-time job, particularly when one considers that
the intelligence officer must also continuously provide his
command - and, in addition, assist in providing subordinate,
adjacent, and senior commands - with the information and
intelligence they require for their day-to-day operations as
distinguished from that needed for estimates. It is for this
reason, rather than any slavish devotion to doctrine that, as
Mr. Smith points out,** some persons hold that the intelli-
? FM30-5 and Principles of Strategic Intelligence, AC of S, 0-2 (Feb.
50).
?? As Smith puts it: "We are told that it is the function of the com-
mander, not of the intelligence officer, to decide what counteraction
to adopt against enemy capabilities and to judge what the success
of such counteraction may be."
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
gence officer should not deal in the capabilities and lines of
action of his own side. Mr. Smith is correct in saying that
some persons oppose this from wrong motives, but that is not a
fault peculiar to the military. It should also be pointed out
that many planners have a supercilious view of intelligence and
intelligence officers. They fancy themselves equally competent
in intelligence matters. Indeed, most of them are, but the
reverse is also true. Most intelligence officers are fully com-
petent planners. Since each has a full-time job, however,
each needs to tend to his own knitting to get the job done well.
There needs to be, and in good commands there is, continuous
close liaison at all levels in the intelligence and plans sections.
Historically it is true that many commanders have leaned as
much or more on their intelligence officers in planning matters
as they have on their planners. In even more cases, after the
whole staff was thoroughly informed about the enemy, the role
of the intelligence officer appeared to be less prominent. It is
noteworthy that this usually occurs on the side that is winning
or has a preponderance of force. When things are tight, the
intelligence officer is in great demand and, we might note, his
neck is way out.
We noted above that the commander also makes an estimate.
His estimate takes the enemy capabilities - presumably as
developed by the intelligence officer - and, in the light of
each capability, studies each line of action open to the com-
mand to determine the one that best accomplishes the mission.
He determines the lines of action open to him by having full
information about his own forces - their position, condition,
morale, supplies, supporting forces available and so on. Just
as the intelligence officer contributes the information about
the enemy, so many other staff officers contribute this other'
information which the commander must have to make a sound
decision.
Let us then keep clearly in mind that, in military usage, they
intelligence estimate sets forth the enemy capabilities. The;
commander, for his part, uses that estimate in conjunction
MORI/HRP THIS PAGE
SECRET 23
with other information (there may be a logistics estimate, an
air estimate, etc.) and makes a final "policy" estimate to deter-
mine the line of action which will best accomplish his mission.
The Military Theory of Capabilities
Many of the difficulties which Mr. Smith points out in the
application of military usage in the field of national policy
stem from the fact that in the national field we do not have
the same common understanding of staff and command func-
tions that obtains in the military. This is true both because
the "staff" in national policy affairs, though to a degree com-
parable, is not a close parallel to a military staff, and because
many of our policy-makers are not experienced in or familiar
with staff functioning.
Against this general background, we can now examine.-Mr..
Smith's advocacy of the concept of "gross" and .`.'net'.' capa-
bilities and his contention that war-gaming should be used to
improve the usefulness of our intelligence.
In reference to the first matter Mr. Smith points out the
need to recognize that enemy capabilities are one thing when
we study them in the light of one of our own actions and quite
different when we consider them in the light of another.
To indicate these differences he uses the expressions "gross
capabilities" and "net capabilities." Use of these terms brings
to mind the idea of a fixed measurable quantity like the gross
income of General Motors and, similarly, that a "net capa-
bility" is like GM's net income. It is quite clear that such a
concept is not accurate.
Pursued to the logical end, gross capabilities would be capa-
bilities, as it were, in a vacuum. Such capabilities have no
practical meaning, both because they are limitless (without
?;)liosition the Soviets can do almost anything) and because
there are no true vacuums in world affairs.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
In a sense capabilities are always "net." But they are fixed
only in reference to one given set of conditions. As these con-
ditions change, the capabilities change. They are a moving
picture, not a still photograph. The Soviet "net capability"
to induce a peripheral war in Thailand is one thing if Thailand
has the political and other support of Burma and the SEATO
states and quite a different thing if it does not have such sup-
port. Indeed, the timing and extent of such support changes
the "net capability." In military usage capabilities are always
what Mr. Smith calls "net." The intelligence officer deter-
mines the enemy's capabilities as of a given time and in the
light of given circumstances.* This idea is readily applicable
in national strategic intelligence.
What Mr. Smith calls gross capabilities could perhaps better
be thought of as "basic" capabilities. For example, intelli-
gence officers can readily estimate that by 1959 the Soviets
could have a stockpile of X hydrogen bombs, Y rounds of atomic
artillery ammunition, Z intercontinental bombers, W army
divisions, and V major naval craft, and could still meet the
industrial requirements of their civilian economy, provided
they give no more than the current level of military aid to
Red China and the Satellites. On the other hand, if they
curtailed production of equipment for the Red Army and
Navy they could contribute more to the armament of China
and the Satellites. These are capabilities. They are basic
capabilities to produce or take general action not normally
subject to interference. Further analysis and research can
develop what, under various assumptions, the Soviets can do
with these resources and thus can determine their capabilities
to act. Perhaps it is this distinction that Mr. Smith has in-
mind when he speaks of "gross" and "net." Even if this is,
the case we would still be loath to accept the concept because,'
in the general sense of the term, even such "gross" capabilities
'See quotations from Dictionary of US Military Terms for Joint
Usage, cited by Mr. Smith; also the description used at the Strategic
MORI/HRP THIS PAGE
are "net." Rather than adopting misleading terms like "gross"
and "net" we seem to be better off if we stick just to "capa-
bilities" and understand it to apply, as in basic military
doctrine, to a stated set of circumstances.
The second point in Mr. Smith's thesis that we wish to
examine is the matter of war-gaming. He laments the fact
that accepted practice frowns on having intelligence officers
war-game the plans of their own side. We do not concede
that this "frowning" is as prohibitively effective as Mr. Smith
contends. To the extent that it does exist, it is directed
against the idea of having the intelligence officer play both
sides. This is logical. The intelligence officer cannot be
-expert" on his own, resources and plans as well as on those
of the enemy. As pointed out earlier, the latter is a full-time
job. To the extent that he thumps for joint war-gaming by
intelligence and plans personnel as a device to assist in im-
proving the usefulness of intelligence estimates, however, Mr.
Smith is emphatically right.
War-gaming for this particular purpose is not used as widely
in the military as it might be. But the concept of war-gaming
for other purposes with all staff elements participating is well
established. It could easily be used in the more complex field
of national estimates.
War-gaming has been modified radically in recent years with
the employment of advanced mathematics and electronic com-
puters. These techniques leave much to be desired in the
military field and many of them could, at the current stage of
development, be used to only a very limited extent in reference
to the "imponderables" of national policy affairs. The more
conventional type of war-gaming, on the other hand, could
certainly be used across the board and with every possibility of
making our intelligence estimates more useful.
Mr. Smith's observation that national policy-makers have a
more complex problem than military leaders is valid, and it
has an important bearing on the activities of the intelligence
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
services which support them. The national policy-maker must
consider a great variety of "capabilities" which interact on each
other. For example, a sociological change in Germany may
have an important repercussion in the political capabilities of
France. Furthermore, it is always difficult to determine the
"facts" in many areas of interest. The military leader usually
knows how many and what kinds of guided missile squadrons,
atomic bombs, fleets, and army troops he and his opponent
have. The political leader is always far less certain about his
"forces" and those of his allies. There is even more uncer-
tainty about the resources the enemy can bring to bear. To
illustrate, we can be sure that Khrushchev's advisers have
many a headache estimating how effective the Satellites and
Communist China really are and what assets the West
actually apply in various situations. In such a field, therefore
there can be no one "net" capability. There are as many "net'
capabilities as there are variant situations. Mr. Smith appears
to think that intelligence officers should compute these "net'
capabilities by their own efforts. It would seem more logical
that they should be worked out in conjunction -and we do
not mean concurrence - with the planners. Intelligence offi-
cers and planners must sit down together and thrash out a
the angles. This is precisely what happens in an efficien
military staff in time of war. The formal estimates of capa
bilities appear only when a radical change in one's own or
the enemy situation takes place. For example, after "The'
Bulge," 21st Army Group conducted an extended and more
or less "conventional" campaign to gain the Rhine. It was
obvious that crossing that formidable obstacle would call for
different types of action and support. An estimate of the
situation was essential.' This, in turn, meant that intelli
gence forecasts and estimates had to be produced. At such,
times a new "stock-taking" is in order. At other times, day
to-day close coordination by the working intelligence officers
planning.
MORI/HRP THIS PAGE
and planners, with a check on interpretations of major impor-
tance by the senior intelligence and plans officers, is the best
modus operandi. It keeps all concerned aware of enemy capa-
bilities applicable to the prevailing conditions.
In the national field, a similar condition could obtain. Un-
happily the lines of demarcation in staff organization are not
as simple and clear as in the military. Instead of overall
planners like those in the Joint Staff or in an international
staff such as the Combined Staff Planners of World War II,
we have political planners in State, military in Defense, eco-
nomic in agencies like OES, propaganda in USIA, etc. Each of
these has some form of intelligence support of its own. These
intelligence agencies are tied together by CIA for national
purposes and planning is brought together in the NSC. How-
ever, there is still a vast amount of "sprawling." Parentheti-
cally, it should be noted that this statement is a description
of a condition; it is not to be construed as an unfavorable
criticism. This is not the occasion for such criticism; and
it is by no means certain that highly centralized planning and
intelligence would be best, or even better, for the country.
Here, we want simply to note that close integration of intelli-
gence into planning is difficult because of the decentralized
planning and operating mechanism in the US government.
A great deal of informal coordination on the working level
clues take place. This is all to the good and should be
encouraged. This complexity of organization and operations
in the national field results in a greater need for formalized
1sstimates and is, in itself, a justification for the use of the
war-gaming principle. However, with all due respect for the
.,kill, wisdom, and judgment of our intelligence community,
we should not leave war-gaming as a basis for decisions to
then alone. The danger here is at least as great as it is to
have the planners do it alone. We have suffered on both the
military and the national plane from an unwillingness (or
inability) to accept and understand available intelligence. We
III ecl not repeat such gross errors.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
With little or no information of our own plans and resources,
the intelligence officer can still tell the, planner what resources
the enemy can have at a future date and the general kinds of
action he can initiate with them. If the commander an
planner want to know what results the enemy can achiev
with these resources and actions, the intelligence officer must
have knowledge of his own resources and plans.
Applying this notion to the current situation, we can expect
national intelligence officers to tell us what resources th
Soviets will have for peripheral wars by 1959 without much
guidance as to our own resources and national plans an
policies. But they can tell us where and with what likeliho
of success the Soviets can use those assets only if they kno
the opposition which the Soviet action is likely to meet. Join
war-gaming would provide such interchange of informatio
It should make for a healthy interplay between intelligent'
and planning and probably result in improving both.
Estimating Enemy Intentions
In Colonel Kirtland's paper we have a more restricted an
therefore more specific subject for consideration. He objet
to what he describes as "unrealistic resistance" to the use
intentions-analysis as opposed to capabilities-analysis in Intel
ligence estimates. He holds that we need to consider bot
By inference, he is most directly concerned with combat in
ligence. He makes clear, however, that his conclusions appl;
to strategic intelligence as well.
After analyzing what Colonel Kirtland has to say, we
agree with his main thesis that both intentions and capabiliti
need to be considered. However, he has not hedged his p
posal with essential safeguards and his arguments ag
the "capabilities doctrine" contain very serious weakness
We will review these arguments and then develop our o
conclusions.
MORI/HRP THIS PAGE
In order to evaluate Colonel Kirtland's contentions, it is
important that we have a common understanding of the
meaning of "the capabilities doctrine." The burden of this
concept is that in a combat intelligence estimate, the intelli-
gence officer should present to the commander his best estimate
of the enemy's capabilities rather than the enemy's intentions.
The doctrine goes further: it holds that the commander in his
estimate should consider each of the lines of action open to
him in the light of each of the enemy capabilities in arriving
at his final decision on a course of action. It is important to
keep in mind that the doctrine has these two aspects: first,
the intelligence officer is to determine capabilities; and second,
the commander should make his decision only after considering
all the capabilities.
An elaboration of this doctrine which is too often forgotten
is that the G-2 is expected to give the commander his con-
clusion as to the relative probability of the exercise of any of
the enemy capabilities, where there is evidence to support such
a conclusion.*
Earlier doctrine had held that the task of the intelligence
officer was to estimate the mission of the enemy and, from that,
deduce the lines of action the enemy might take and then to
determine their effect on the courses open to his own side. This
doctrine invited a refined form of guessing as to the enemy
mission and encouraged consideration of intentions in the
deduction of enemy lines of action.
The new capabilities doctrine was developed after World
War I because it was felt that earlier doctrine introduced too
much clairvoyance into military problem-solving (which is
what decision-making really is), and that it came too near
urging officers to guess the worst the enemy could do and to
,;take everything on that. It was believed that the "capa-
lnlities" system was more "scientific" and more nearly in
accord with the facts of life. This conviction was illustrated
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
at the Command and General Staff School just b
f
e
ore Worli
, War II, when one of the instructors "clinched" the argumen
in favor of b
asing estimatb
es on capailities by showing tha
in World Wa
I
r
von Kluck hadhd
cange his mind four tim
in one day and actually issued three different orders.
A concomitant of the acceptance of the capabilities doctrin
one wh
an
d
y
o a
voc
e,
basing estimates on enemy intentions just hasn't been brough
nA,,..__- ~.
un nrnnarlcr T
o
come to be considered the equal of advocating mind-readiin
or th
e use of a ouija board. Advocates of intentions-analys
like Colonel Ki
tl
r
and object m t thi
oreos anti-intentions prej
dice than to th
e capabilities doti
crne per se.
In marshaling support for the thesis that our doctri
ne ne
review and, in particular, needs to give more consid
ti
era
on
intentions, the critics tend to make some amazing misinterpr
tations and t
o neglect some cil ft
rucaacs. We agree that o
doct
i
r
ne needs recasti bt
ngu we must, in fairness, keep
record acc
t
ura
e and logical
.
Colonel Kirtland's objection to current doctrine i
s
,,
on three main points: first, "a nation or a comma
d
n
er m
w
have a preponderance of force if he bases his decisions o:
conservative"; and third, the enemy's potential capabilities ai
t
no
adequately considered.* We will examine each of th
points in some detail.
The statement that the capabilities doctrine is useable onl
when you have a preponderance of force is clearly erroneo
It
is a very practibl dt
caeocrine when you are on the defens
and even when you are the hunted in a pursuit. To ho.
th
o
erwise is like saying you cannot use the principles of are
? The third point is paraphrased because the actual statement is n
very precise. However, subsequent explanation makes clear that
means what h--.. . - -
MORI/HRP THIS PAGE
metic when you are in debt. The capabilities doctrine - and,
for that matter, any other doctrine - gives you a discouraging
picture in such cases, but that is the picture you must face.
In an adverse situation, the doctrine is designed to indicate
which line of action would have the least adverse result. In
other words, it indicates the course of action which would get
your nose least bloody.
The second criticism, that application of the doctrine gen-
crally results in conservative action, is to a large extent true;
but it is true because, in matters of life and death, leaders
generally tend to be conservative. Usually they should be.
The criticism is justified only to the extent that the going
doctrine makes it easier for leaders to be conservative. This
is particularly true when officers take the view which an
allegedly bright and "successful" officer (he later got a star)
expressed when he said: "I teach my officers to select the
line of action which gives them the best chance against what
tliey figure is the enemy's most dangerous capability."
It is this use of the capabilities doctrine that brings on the
,-ritirism of conservatism. Actually it is a reversion to the
elder doctrine. It is, in fact, a form of intentions-analysis
because the user assumes that the enemy will exercise a given
capability. Such use does not condemn the doctrine itself,
any more than the fact that some men get drunk justifies the
rondemnation of all whiskey. Current doctrine holds that the
commander shall select the course of action which, in the
.`are of all the estimated enemy capabilities, insures the most
c !7ecf ire accomplishment of the mission. This is not the same
thing as saying that he should select the one that gives the
greatest certain.ty~ of accomplishing the mission. Cleariy the
.
most certain course might be the most bloody while a slightly
more risky line of action would be less costly and might accom-
plish the mission in a shorter time or have some other advan-
tage. The selection of a line of action requires a balancing
of costs and gains under the various possibilities. It also calls
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
.
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
for what is known as "military character." No matter wheth
_.
we use capabilities or intentions, the decisions will reflect tha
The third argument is that use of the doctrine prevents con
sideration of potential capabilities, meaning those that develo
between the time the estimate is made and the action tak
place. This, of course, is woven of the very flimsiest clot
The doctrine is based on the use of capabilities which tk
enemy will have at the time of the action for which one
planning - not the capabilities at the time the decision
made. It is the capabilities forecast for the action-time
.
one accepts the argument, he must also accept the conch, sio
,
that if intentions were used in the analysis, one could no
use forecasts of intentions. On this score, then, one woul
be as badly off under one system as under the other.
One other serious error in Colonel Kirtland's paper that
must bring out is the failure to show that Army doctrine h
for years made clear that in strategic intelligence-as
tinguished from combat intelligence - both intentions an
capabilities are considered. Official doctrine and teaching a
the Strategic Intelligence School and at Army schools ha
emphasized this point at least since World War II.
The Role of Intentions in Intelligence Estimates
So far we have been concerned with showing that the
ments presented against the capabilities doctrine are not v
good or conclusive. This is not the same as saying that,
are trying to build a case against intentions-analysis. Ac,
ally, we do not intend to do so. We will weasel but, we belil
with good reason. We agree that use should be made of bo
capabilities and intentions in developing estimates, but we Ii
that one must be equally objective and "scientific" in de
mining either of them.
MORI/HRP THIS PAGE
Having noted that the common arguments against the capa-
bilities concept are not too decisive let us note a few of the
weaknesses of that system and indicate some of the strengths
of the intentions approach.
The faults of the capability system are two-fold. First it
tends, as Colonel Kirtland points out, to cause intelligence
officers to include remote possibilities as capabilities. They
forget that the doctrine calls for the consideration of only
those capabilities which bear on the accomplishment of one's
own mission. Second, and despite strong language to the
contrary in Army training, the doctrine seems to justify lazy
intelligence officers to feel that they have done their bit when
thev have made one- forecast of capabilities. This is most
unfortunate. Intelligence officers must keep capabilities under
continuing study to narrow them down. For example, in
September of 1943 the predicted capabilities of the Germans
vis-a-vis the Normandy landings were of a given order. As
time went on, the Allies developed certain techniques and
equipment and new forces became available. On the Axis
side, Italy was knocked out of the war, and the Germans
committed some of their forces in new areas. Consequently,
the enemy capabilities changed continuously so that by June
1944 they were far more limited than could possibly have been
predicted in September 1943. SHAEF intelligence kept a con-
tinuous spotlight on these capabilities during this period. So
it should be in all operations. The good intelligence officer
keeps on the ball as long as there is time to influence his own
side's line of action. In many cases the situation develops so
that at a point the enemy has only one capability. This hap-
lxened at Falaise and in the Ruhr. Eventually, the Germans
could no longer disengage their forces. They had to stay and
fl~,lht. This idea was also illustrated in General Eisenhower's
statement to the effect that after a given time he could no
banger influence the course of the Juggernaut that became
the Normandy assault. For a considerable period he had only
one capability. Just how long the German G-2 was useful
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
by keeping tabs on that has not been made clear. Our teat
ing does not emphasize this concept as clearly and firmly as
should.
As we have already noted, World War I provided a startlingi
effective case to bolster the capabilities doctrine. Similarl
the Civil War and World War II give us particularly fine cas
for defense of intentions-analysis. In the Civil War, oppos
commanders often knew each other personally. They used
knowledge in their planning. They knew the training, abi
_
ties, and personalities of their opponents and, hence, coul
determine the line of action the enemy was most likely to tak
In a sense, of course, this too is an assessment of capabiliti
but there is no point in splitting any unnecessary hairs.
ordinary language, such an evaluation results in a predictio
of intentions. There is a grey zone where capabilities slid
into intentions, but for our purposes, we will lean to the co
servative side and call the borderline cases intentions.
The World War II support for intentions-analysis is in so
ways even stronger. It stems from the fact that the Japan
M
tendency`to fight to the death was so effectively ingrained that.
to a very marked degree, capabilities to take other lines
action were not meaningful. To a lesser extent this s
situation applied in the European war where Hitlerism mold
capabilities.
One can make a very good case for the contention t
enemy intentions should properly be considered under tl
capabilities doctrine because they are a factor in the comb
efficiency of the enemy. To accept such an interpretati
without clearly labeling it, however, would simply be a wa
of getting around the intent of the doctrine and have tli
disadvantage of not calling intentions by their true name.
Experience in all walks of life shows clearly that a failur
to make a thorough study of one's opponent to determin
his motivations and his mental and psychological reactio
as a basis for estimating his future action is worse th
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
unwise. The press is full of stories that the USSR is very
active in this field and has attained great successes, perhaps
as a concomitant of progress in brain-washing and psycho-
logical matters generally. In our zeal to make sure that
training will make commanders and intelligence officers "ob-
jective" and "scientific," we may have gone so far that we
have tended to overlook the obvious. Certainly, the mental
makeup and attitude of the enemy is as much a "fact" as is
his training, his morale, his organization, or his weapons.
Surely then it is logical to consider intentions. Equally surely,
it is important to do so objectively and to know what you are
doing. If you are an intelligence officer, it is most important
that you alert your chief to the fact that you are considering
intentions.
In the discussion so far we have used examples and applica-
tions in the purely military field. The conclusions are valid
in national intelligence as well. In fact, intentions of a nation
or a government can be determined with more accuracy than
those of an individual commander. These intentions are
shaped by many clearly observable facts such as past actions,
sociological conditions, cultural characteristics, internal politi-
cal pressures, economic circumstances, and a host of others.
The British exploited their understanding of German inten-
tions in both World Wars and it was not uncommon to hear
their intelligence officers use such expressions as: "the Hun
Is sure to - - - -," and "the German probably appreciates."
They personified the enemy government and high command.
On the other hand, the Germans seem consistently to have
missed the boat. They clearly either did not or could not
evaluate US and Russian national intentions properly in either
of the World Wars. The evaluation of national intentions in-
volves a more comprehensive field of thought than does the
evaluation of the intentions of an enemy commander. How-
ever, the task is no more difficult. Even if it is, it must be done
because the rewards for success and the costs of failure are too
great to permit neglecting the job.
MORI/HRP THIS PAGE
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Where does all this get us? It seems is i
di
n
,
cate that
Colonel Kirtland says, a proper doctrine would be to inclu
both ca
bilit
pa
ies and intti i
enonsn all estimates as we now
in th
t
e s
rategic estit H
mae.owever, we should expand
princi
le
p
will impress on all .concerned that they need't
i
o
apply the m
e
rigid tests to all evidence bearing on intentions and that co
cIusions bas
d
e
upon themllh
ceary sow that this is the
Since all concepts and doctrines wind up in a "form" of so
enrf - -1-4.
_ ,.
as
In the military field the solution is eas
. All
l
y
we
need to do
i
the commander's estimate * is to insert a paragraph on "enem
tntPntinne- mi-
statement, either to the effect that there a
n
re
o reliai
indications of enemy intentions or that certain stated e
dente i
di
t
n
ca
es an intti t
enono exercise one or more of th
ca
abiliti
p
es
.
In the intelligence estimate, we need merely insert ti>a
"cnmhat G{~ ninr ..11
characteristics which shape or have a major infl
on
uenc
ru
actions. In addition, we should add a paragraph on enem
estimate. This one should also present the ucriticallle der
upon which th
e estimatef itti
onenons is founded.
Such a detailed analysis of combat intelli
ence d
t
g
oc
rine
warranted at this juncture because, as Mr. Smith points o
so much of the concept and procedure of
. .
int
process. The additions to military command and intellllge
estimates which we have
Vt~
r
p
oposed here could b
e paralleled our training for nati
onal strategic itlli
negence.
Our current doctrine probably goes too far in playing d
intentions-analysis. Going all out the
+,,e_ ..
o
MORI/HRP THIS PAGE
iuly be worse. It would encourage clairvoyance and, in
ddition, might discourage the continuous effort to seek for
ew indications of capabilities. The stress on measurable
hysical facts is justified. While we are making important
strides in understanding and measuring motivation and mental
processes, we are not yet far enough along in that field to
measure intentions as precisely as we can capabilities and, as
Colonel Kirtland notes, the danger of deception is a very real
one. Even so, since decision-making is so inevitably bound
^p with consideration of the personal element, it is the better
part of discretion, and of valor as well, to consider intentions.
They are so often the sparkplugs of human action.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
SECRET 39
EnrroRs NOTE: Whenever books or articles ap-
pear that have a close relation to the subject of a
monograph, we plan to include a Bibliographic
Section. This will have the primary purpose of
directing the reader's attention to items in the
existing literature, overt and classified, which in
our judgment make a contribution to the devel-
opment of sound intelligence doctrine. We think
the following is one such item.
(bt. Sanford H. Kirtland, Jr., "The Hazards and Advantages of
Estimates of Enemy Intentions." Thesis, Air
War College, Air University, Maxwell AFB, April
1954. Mss. CONTF. 50pp.
also summary in Air Intelligence Digest, January
1955.
In this paper, Col. Kirtland comes frankly and vigorously to
imps with the caveat in traditional military intelligence doe-
-ine against estimating enemy intentions-or, to put it
a:11 ither way, against breaking down the distinction between
:'.emy capabilities and enemy intentions. Col. Kirtland is
'xr from contemptuous of this doctrine; indeed, he makes an
client case for it, emphasizing the dangers of second-
-ssing and of assuming that the enemy will choose to do
rrtty much what a US commander would do, in a similar
'.'?a lion. He emphasizes, too, the danger of writing up an
mate of the Situation from even the shrewdest guess of
my intentions, thus inviting disaster if the guess turns out
,c shrewd but wrong. In brief, this thesis is no hatchet job.
MORI/CDF THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
40 SECRET
What it is, on the other hand, is a most sensible investiga
tion of the traditional doctrine and an invitation not to f
into a variety of naive traps where the estimative process
concerned. First of all, the author points out that the distin
tion between capabilities and intentions is sometimes synthetic
The line can be more easily drawn in the abstract than it
can in real situations - especially, we might point out, in situ
ations that count the most, when a US commander has to
spread out thin resources to meet a variety of possible enemy,
moves. Any intelligence officer (as Mr. Smith argues above?1
obviously works from estimates of intentions in that he ex?
eludes from his situation-estimate a whole series of outlandish
and, from the enemy point of view, self-defeating gross capa
bilities. If the clear enemy objective is to seize a piece of land]
it is not very instructive to point out that he is capable of ail
immediate, orderly retreat.
A
Second, according to Col. Kirtland, the intelligence officer
is forced into estimating intentions (or probable courses o[
action) precisely because the US is no longer in a position of
undoubted preponderant power from which it can prepaie
for and can thwart any and all enemy capabilities. Which is
to suggest that the traditional doctrine is outdated. As Mr.
Smith says:
There has been much debate, among the military, on
whether an intelligence officer should presume to
put into his formal estimate an opinion as to which
of the enemy capabilities listed is most likely to be
implemented ... Some have even held that the com-
mander himself must not make it, but must treat all
enemy capabilities as if they were sure to be carried
through, and must prepare to deal with them all
This latter doctrine is somewhat academic. (Emphasis
added.)
Col. Kirtland and Mr. Smith both seem to be saying that
these days the intelligence officer may pay lip service to the
traditional military doctrine - may insist that he is follOW6
MORI/CDF THIS PAGE
SECRET 41
ing the book on the distinction between capabilities and
intentions - but cannot possibly keep the distinction clear
in practice.
Finally, the author concludes that there is no inherent
drawback in estimating intentions: to do so with reliability
simply puts the burden on finer judgment, on better back-
ground and training, and on better personnel selection of
estimators. He might also have added that since estimating
intentions is what the intelligence officer in fact does, some
of the time at least, it would be well that he do it consciously.
The real danger is that the estimator might think he is dealing
with relatively sure and scientific capabilities data (claiming
:dative certainty for his conclusions, therefore) rather than
with speculative premises about enemy intentions.
Col. Kirtland is writing, of course, strictly about military
:ntclligence. But most of what he says can be translated into
frame of reference of the civilian intelligence agency - as
1t r. Smith's paper demonstrates -with some valuable instruc-
t:on for all of us. This is, at the least, a thoughtful contribu-
-en to the subject.
MORI/CDF THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
MORI/CDF THIS PAGE
MORI/CDF THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
pto a pressed in the Sfud~e '
offiga e
utha T eyooepresent
.tin intelligence Agency o of thee~ ce o i
` ... erecting" the
atrial contiatns information
efense f the United States within$'the$meanin~
;pianage laws, Title 18 ;USC; Secs 793 and 794; tlie;
fission orevelation of which to an nnauthoed Pe
rabbi} by law: *4 S..,JS ' a
'yp Y t ~f + X 1 d '~ '~" 3+ ~<
Subs ", a issues will be disssemmated1wlde y' o%g,
gegcy To make,~sure~cfrecen~ng }~~a
44ra~copes! pfieasettwll~tf extensionk
EDITORS INTRODUCTION
BACK in 1951 when the Office of Research and Reports was
just being set up, the then Assistant Director, Max F. Mil-
likan, wrote a long paper for his staff on the functions and
methods of economic intelligence. In it, he spelled out the
reasons for ORR's existence, its major and minor tasks, and
the methods by which these tasks might best be performed.
The paper was, in effect, marching orders for a new organiza-
tion as it embarked on its job.
We have here printed the greater part of Dr. Millikan's paper,
without any substantial changes and without any attempt to
bring it "up-to-date." Why? First of all, not because Dr.
Mtil,ikan was the first AD/RR; and second, not because the
pa )er was an official document defining ORR's operating prin-
?ples. Indeed, recalling that Studies in Intelligence will pub-
:..sh only unofficial, individual contributions to basic intelligence
,ioctrine, we are printing Dr. Millikan's paper in some sense
respite these facts about its author and its original purpose.
our reason for publishing the paper is this: it is, we feel, a
d:stinguished contribution to the study of intelligence analysis
methodology. And its application is by no means limited to
economic intelligence; the same order of analytic problems, the
ume problems of sources, extent of information, competing
:!-quirements, liaison and coordination arise in any intelli-
o nce activity. The same problem Dr. Millikan addresses,
t:at of building authoritative knowledge out of fragmentary
-.uces, is perhaps the central problem of the intelligence
;:( ess as a whole.
?here is no need to rehearse in detail Dr. Millikan's quali-
,tions for tackling such problems. He was for several years
-' nior official of this Agency and is now Director of the Center
. International Studies at M.I.T., an organization that has
',nda.cted important research on Soviet and World Communist
41"w -s since its inception in 1951.
MORI/CDF THIS MORI/HRP THIS PAGE
PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
ii SECRET
Howerton's paper complements Dr. Millikan's, on
Mr
.
levels. On the level of economic intelligence research, it give
l
yz
a current view of the methods devised since 1951 for ana
in successive approximations to the complete picture,
thus, in effect, a case-study of
t economy; it is
i
S
,
ov
e
d of successive approximations described in broad tei
th
o
me
by Dr. Millikan. On a more general level, Mr. Howerto
of racani
rovides a s
_- . -.,
paper p
ues available to the intelligence analyst and of
i
h
n
q
tec
potentialities of overt (or, at least, easily obtainable) inf,
t to An with 't
h
a
mation - if only the analyst knows w
resources available to him.
Mr. Howerton joined this Agency in 1951 afteTTa career
nd
re
e
a
r-
co po a
a unique combined competence in chemistry, mathemat=i
e of natio*
and languages and is fellow of more than a scor
Hower
M
r.
and international professional associations.
lli
t
g
e
is, at present, Assistant to the Deputy Director/In
(Planning) .
In this issue we also inaugurate a continuing featurl
d Common D
-r---
Notes an
l
11
ber of letters on the first two issues - many, in themset
tions
t
ib
---------
r
u
Substantial con
ciples and methods. Of these, we here publish two, both
ana
the
to
i
e
"
C
r
v
and respons
the subject of "capabilities
-1-
Studie
Into
+ ?.-
issue of
is Director of Intelligence, U. S. Air Force; Mr. Alan J. P-
is JIC (London) Representative to this Agency. We are
l to both for permitting us to publish their comments.
f
u
we very much hope that, as subsequent issues appear, we
estii1l
u
d
s
gg
keep on receiving your coiiments, criticisms, an
ou th
what
k
y
now
publishable or not. We would like to
the series.
MORI/HRP THIS PAGE
THE NATURE AND METHODS OF
ECONOMIC INTELLIGENCE
by I-lax F. Millikan
D URING the first half of 1951, ORR was engaged in taking
an inventory of its ignorance concerning the economy
of the Soviet Bloc. The main purpose of this inventory
u::s to establish a basis for planning a program of basic
research to which ORR should address itself. Such a program
mist spring from a clear conception of why the US Govern-
nnwnt needs foreign economic intelligence, what foreign eco-
nomic intelligence is, what role ORR should play in the total
,i-onomic intelligence effort, and how the peculiar character
,,, the Soviet economy and of our information about it influ-
cnces the methods that we use. This introduction is devoted
to some comments on these four topics.
4V hv does the solution of our national security problems depend
in part upon adequate foreign economic intelligence?
Foreign economic intelligence serves at least five purposes
the design of policies to preserve our national security.
T!,cse five purposes, which should be kept continuously in
in planning our economic research program, are as
..Ows:
To estimate the magnitude of possible present or future
or other threats tovur52iveS and our allies. A pv~2n-
a' enemy can undertake successfully only those military
,,; y : at ions which its economy is capable of sustaining. In the
- ' Y short run, its strength may be measured in terms of the
::-%po%k-er which it can mobilize and the stocks of finished
";L;x,ns of war and military supplies which it has on hand.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
SECRET 3
SECRET
however, military potential f
Increasingly in modern times, has come to depend upo11
aigns
m
p
inudin
anything but the briefest ca
,,, ces available to a nation,
es
well
civilian economy
those necessary to support, the instruments
sary to produce and operate the of w
neces
picture of the magnitude of the present and pons
A clear p' guide us as to 1
t is needed to
threa
future military or other over-all magnitude of the defense effort in which we m
ur freedoms in the event of a
reserve o
engage in order to p
d location of possible pr
acter an
2. To estimate the char or other threats. Decisions which the USS
or future military make with regard to how tl
or any other potential enemy
will allocate their resources limit litanthey can choose to
y installations in
If they elect to invest lag y corr
Far East, their potential for attack in Europeri their in# of j
d This 16 no li mitdn
ingly restricte. 'hatations are place o
i
n
ns
ng decisio
tions but rather of see
in the future by
to them
courses of action open lu onnrat,inn of their total resou
they mare wu?y the character
A principal purpose of thus estimating possibly
r own defense effort so that ii
g ou
guide us in designin is tog dangers
protect us against real rather than imaginary
e of the
in
g, --- potential en{
3. To assist us in es-Lima
ble, the intentions of the USSR or and their presentd
of
bution permit him to select, any
_ range certain
xxn+-hin this
f nr&
probable coursesu indicat ions as to which alte i n0mic events may __.,_ an,~ where and when.
its
-
----
the Sov
tions may be veryimp01
f inten
rJ u +n mee,
Ainat'ons o~ These ul~, w r re ara,l0
a
us to adjust our defense pp
assisting
most probable -0- 4. To help policy-maker s decide what we can do to
?
r other threats by imps
o
possible or probable military them out. This
enemy's economic capabilities to carry
SECRET
MORI/HRP THIS PAGE
measures that can be taken to weaken him in advance of hos-
tilities and thus delay or prevent his decision to engage in
them, as well as measures to weaken or destroy the economic
basis of his military power should he choose to commit it in
general war.
Economic intelligence can help in suggesting such measures,
in estimating their effectiveness, and in forecasting the enemy's
probable reaction to them.
5. To assist in estimating the probable development of the
relative strengths of the East and West over the next few
rears if global hostilities are avoided. A major purpose of
these comparisons is to guide US policy-makers. The preced-
ing four objectives are concerned with steps which the United
States can take to defend itself against actions of a hostile
power. Equally important is the design of that political policy
which will have the' best chance of achieving our objectives
without hostilities. Essential to the planning of such a policy
is the most accurate estimate possible of the relative economic
strengths of both sides. There are equally grave dangers in a
serious underestimate and in a serious overestimate of future
Soviet economic strength. Either will produce policies more
likely to bring on war than will an accurate estimate. The
evaluation of Soviet strength implicit in various of the pro-
posals for US policy now being advanced in this country varies
widely from great economic weaknesses to very considerable
economic power. A prime goal of authoritative economic intel-
ligence is to provide the information that will narrow the
"guess area."
% 'hat is economic intelligence?
Briefly, economic intelligence is intelligence relating to the
basic productive resources of an area or political unit, the goals
and objectives which those in control of the resources wish
then; +w serve, and the ways rn wu &Li and the effi l trueness with
which these resources are in fact allocated in the service of
these various goals. There are a number of confusions as to
the nature and limits of economic intelligence which call for
cluiflcation.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
there is sometimes a tendency to reg
In the first place
,
the whole of economic intelligence as encompassed in a mi
begin to be employed broadly in production, they bee
and development phases. When these techniques and o ,
,
draw. In very rough terms, scientific in a ig
precise lines between the segment's are r...pos
"op follon
t 11'
,J s..- -- ---- - -
of dividing it are somewhat arbitrary analytic invert
4 m C
Ano er p
91
leaves off and political, military, and scientific mtel
th roblem relates to where economic intellil
as with the physical inventory of the resources
serve, and the ways in which they are related to ea h;,
coal to operate the railroads. Thus economic m e g
y
a
e
by the steel required for the machinery necessary to mine
t11 ence+
steel from steel plants to tank plants, or, more rem,
er
available to make the rails and the freight cars nee
from which to manufacture the tanks but also by then
plex interrelations in balance. Thus tank production,
, - ,_-- 41.... .,,.-;1shill+4 nf
p
capacities of the economy may be limited less by the ove
4.. 1...nn o11 1-ha r
corn lex web of interconnections among its various parts.
Furthermore, a modern economy is characterized by a hig
as of low or negligible priority.
steel must be allocated to many uses which the Soviets re
g
so
its minimum goals, even in a time of crisis like the preset'
hly misleading. For the United States to achi
is hi
about capabilities, vulnerabilities, or intentions. The Allli
Powers have a total steel capacity which is more than. fo
which they are designed to serve or of the methods employ
4,.11 ....U
m
but only a part of the total economic problem. An invents
struments of production. This inventory is a necessary p
inventory of available resources of labor, raw materials;
MORI/HRP THIS PAGE
SECRET 5
province of economic intelligence. Military intelligence is con-
cerned with the character and capacities of the military estab-
lishments of foreign countries and with foreign targets for our
own military efforts. Where the character of the military
establishment depends upon rates of production or where
the target of our military effort is the economy of the poten-
tial enemy, the lines between military and economic become
blurred. The output of final military equipment and the
physical targets on which our military forces must concen-
trate are clearly a prime concern of military intelligence. On
the other hand, economic analysis is required to portray the
complex nexus of economic support on which military produc-
tion depends and to pursue the economic chain reactions which
might result from the destruction of particular producing
facilities.
The overlapping between political and economic intelligence
is even greater. One of the best ways of studying the goals
which a collectivized state wishes its economy to serve is to
examine the institutional machinery that it establishes to
guide economic processes. Thus certain of the institutions of
government, although in a sense political phenomena, may
have profound economic significance. On the other hand, eco-
nomic conditions are of course an important determinant of
the attitudes, loyalties, and composition of politically impor-
tant groups. In these borderline areas, it is the purpose and
object of investigation rather than the disciplines employed
that determine whether intelligence is properly to be termed
economic or political.
A final point of importance which the analyst must keep in.
mind is that economic intelligence is not always the same
thing as economic information. Even the most basic economic
intelligence should always be produced in relation to the
t>otds of some intelligence consumer. The Central Intelligence
Agency is charged with producing foreign economic intelligence
Mating to the national security, and the consumers of its
product are those US Government officials charged with guard-
ft the national security. A vast amount of information -
$ d ed, almost all information - about foreign economies may
MORI/HRP THIS
PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
6 SECRET
nt to national security problems, but it is not econo
l
eva
be re
intelligence until its relevance to those problems is made el
ed
e
know
g
l
It is the function of intelligence not to pursue
o
its UWll bane Nut .,-?__-_
quences of present or future action. Though the intelligeA
__.- + +~?+lr etrivp to kA
o
a
no
1.
t
analyst is
--"J in mind the relevance of information to policy problems, wh{~
alone can transform information into intelligence.
What is the role of ORR in foreign economic intelligence?
Many US Government agencies are engaged in the prodii
The
enc
lli
t
e.
g
e
tion and collection of foreign economic in
line, We mow....
f the foregoing statement of the purposes and na
i
s o
bas
ence alone. We must also consider how,
intelli
i
c g
g
of econom
activities can be made to reinforce rather than to duplicate
they
i
n
great amount of work which others must carry on
__
ent survey of for
rec
Luial6c; J. --
economic intelligence throughout the US Government sug
ti'
focus of our ac
a number of conclusions as to what the 11
should be.
First, our survey revealed that one of the most urgent
Go - -
th
e
of
e.....?---- -- economic intelligence collected and produced throughou
In recognition of this need the National,
s
i
ssue
security
. rity Council has directed that the Central Intelligence Ag
Although this
i
on.
shall perform this coordinating funct
is directed at our production program, our plans for i
nt of
gence production within ORB must take full accou
with ours
n
l
g
o
coordinating responsibilities which go a
tive effort.
A second conclusion of our survey has been that th
lli
P'
t
g
e
conomic in
most in need of substantial additional e
o
'
I
n
r
because the
fort is the Soviet Bloc. This is partly
o iet
has made awcoo --
partly because the Soviet economic potential is perha',
MORI/HRP THIS PAGE
most critical key to our national security, and partly because,
for a variety of reasons, the economic potential of other areas
crucial for our national security, such as Western Europe, has
been much more extensively studied. The mature economies of
Western Europe have long been an object of study by both
academic and governmental economists. The European Re-
covery Program has stimulated intensive analysis of the char-
acteristics, needs, and prospects of the Marshall Plan coun-
tries. Thus, the economic research effort in man-hours directed
at the USSR and its Satellites has been vastly less than that
applied to Western Europe, although, because of the Iron
Curtain, the effort required to produce comparable under-
standing is many times greater. For these reasons, we have
concluded that the principal effort of ORR in intelligence
production must be focused for the immediate future on the
economic problems of the Soviet Bloc.*
We began this research effort with an inventory of our
knowledge of the USSR itself. This, of course, is only a part
of the problem. The economies of the European Satellites,
whose analysis was our second task, are likewise crucial to the
Soviet economic potential. Recent events have highlighted
the importance of China to our estimates .of Soviet strength
and intentions. A final source of Soviet strength, which must
be another object of our efforts, is the resources that the USSR
could draw upon either now or as a consequence of future
developments outside the present boundaries of the Bloc.
A final weakness of the intelligence effort as revealed by our
Inventory is that the demands which have been placed on the
limited number of analysts working on the Soviet economy
have been so frequent and insistent that analysts have had
little or no time to do the basic research necessary to supply
an.. veers in a confident and authoritative form. If our effort
l1 to be useful at all, it must be on a sufficient scale and of
` The Soviet Bloc excludes Yugoslavia and Finland and includes the
%uropean Satellites (East Germany, East Austria, Poland, Czecho-
ciovakia, Rumania, Hungary, Bulgaria, and Albania) and the
t: astern Satellites (Communist China and Communist Korea).
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
SECRET
much firmer factual foundati
sufficient depth to provide a
for the estimating process than economic intelligence has b
able to produce in the past.
Peculiarities of economic intelligence concerning
the Soviet Bloc
There are a number of special characteristics Of the
wa~sl
e in imp
ha
p
economic intelligence problem which s
are n) h
it
d
t
o
bese
y
u
the methods that can be used to s
. o
diffic
more
k
o
he p
e
ever, all characteristics which ma
r first some 011,
r
a
than that of other areas. We m
y our probthings about the Soviet economy which simplify
factors which make it dif le
_f s
he
w.
t economy is centrally planned
i
e
The fact that the Sov
achieve the goals of a small group of men acting collecti>r
__,-. s? +Hc free, economy Of
facilitates analysis euuLwvu~.,. -
United States the tastes and desiresfl~ P 160 on million e s
ha
unpredictable people al
occurs. The behavior of major sectors f countless consum, 1 _Q
greatly affected by the individual p uantita'
hat unpr
each with a different and somew
plans which doming set of weight. In the USSR there is one
-advertence that anything
b
'
y
all others. Thus it is only
iew of the master pla
f
v
occur which, from the point o
irrelevant or unimportant. This makes the second jog
the elu
b e
namely
economic intelligence described a
,
-
tion of the goals and objectives reat deal easier.
resources wish them to serve - a some clue.
anything that happens can give us
A related point, true to some extent offrvevery econom~~
thing d P nd'
h
e -
especially true of t
eise. The ir~tercenner_.tedness of the econom
everything
its subservience to the master plan mean that there are
~r%,inmic fact can be ascert
different ways in which e timated directly from evidences
Steel production can be
the location and capacity of steel mills or indirectly frog
dente of the manpower employed and of the iron ore:'
11
SECRET
MORI/HRP THIS PAGE
SECRET 9
or alloy metals or other inputs available, from the total output
of all the products made with steel, or from the capacities of
transportation facilities serving the steel industry. The lack
of direct evidence on some of the things that we most want to
know, as revealed in the results of our inventory of ignorance,
emphasizes the very great importance of giving priority to the
interrelations of the parts of the economy. Thus the third task
of economic intelligence, to explain all the complex ways in
which resources are in fact allocated to various uses, is peculiar-
lv essential to building a consistent picture of the Soviet economy.
A third fact that shapes our methods is that technology and
the laws of nature are no respecters of iron curtains. The
Soviets do many things differently from the way in which we
do them, but in many other things they have no choice but
to follow the only industrial technique that exists. Thus the
electrolytic process which produces sodium hydroxide and
chlorine inevitably produces them in the same ratio in the
USSR as in the United States. We can learn many of the
technical limitations on what they are able to do from a study
of US industrial practices. But this must be done with care,
since we know that in some cases the Soviets appear to be
incapable of applying our techniques even where they know
about them, whereas in other cases they have devised superior
methods. Nevertheless, with appropriate caution, useful first
approximations can be reached by the comparative method.
One implication of this for research plans is that there must
br present in our work a much heavier dose of technical and
rngineering thinking than is customary in economic studies.
A characteristic which has advantages and disadvantages
it that prices, markets, and money flows, the stock in trade
o! much economic analysis, have limited meaning in the USSR.
We are spared the uncertainties of the capitalist business cycle,
and monetary dislocations are of little significance. On the
other hand, we are largely denied the benefits of money as a
rrmmmon measure of otherwise incommensurable activities.
14ast of our thinking must be not in terms of rubles but of tons
P:d bushels and bales, of numbers of machines of innumerable
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
10 SECRET
of car-miles, kilowatts per hour, and the
ds
t ki
n
differen
, To add all these things up to an index of capabilities, we m
k of n o s'
t
s
as
concoct our own common measuring rod, a
complexity.
On the negative side is the obvious fact that informati
ited inch
li
m
currently coming out of the Soviet Bloc is very
t
? his does no
ted. Raall
our knowledge is inevitably correspondingly limit+hp Rath,
es do t e
ng
economic cha deal more ab'
tion on earlier periods is a good
and informa
dant. Piecing this together with what we^arengetting f1
e ing
..
exercising som
known about the unknown (through the MiQq
ec
d
..
-a
economy), an
formation through the channels available to.
f i
n
pieces o
ut together a surprisingly reliable ;p
to
ibl
p
e
it is poss
ture. What the scarcity of current informationnmeea s s
nim
ndo n
are co
that we ?e` - -o
but rather that to find out what we need^to^know takes ng
s of pa
h
our
many more
than would
interpretation, and of fitting and adjusting a
ei
Thcum
om
e
d
con
y.} p
necessary in the study of an open
of r
sion i
_ -
tion of his wu~ru
time required which were compiled by the various di
during the course of the inventory-
racteristic of the Soviet problem is that
_i,
nal a
d difficulties of collecting information,
t
s an
the cos
of
more time and thought must be devoted to Bete+rmining
l
rrriv.=?? --__ ..
pieces of additiona
we could secure them. This point should not be overe
r
disc/
t
y
o
sized. As the inven
to give the answers that we need about a good many sub
MIS
ashington.
`J4T
i
l
n
e
availab
is believed to be largely
m
!
cases, what is needed is principally much mu-no
e
hnwev
of a rather row-grain. -
collection appears to be the only way of filling in certain?
-nlrl nnrTn y
p
in
d
t
g w..
y
u
gaps. III s
expected to'
for much more information than one expo
MORI/HRP THIS PAGE
SECRET 11
then sort out the useful parts when it came. When the cost
of information in money and lives is high, however, much more
careful consideration must be given to which pieces of informa-
tion are the vital ones. One of the principal responsibilities
of ORR is to give this kind of guidance to the information
collecting agencies.
The considerations set forth in this introduction do not
determine the details of our research program or of our method
of tackling it, but they do provide a framework of ideas within
which the research program may be carried forward. The next
task is to spell out method and content somewhat more
precisely.
The dilemma of the clamorous customer versus the basic study
The central question of how we should allocate our time
has already been referred to. The problems to whose solution
we are asked to contribute are very urgent. Events will not
wait for the orderly, patient, exhaustive research which alone
Can give satisfactory answers to these problems. If we were
to devote ourselves exclusively to amassing all the facts we
nerd. we would have to tell harried policy-makers that we
could be glad to advise them-beginning in about 2 years.
We neither should nor can stay in an ivory tower that long.
Earn if it were possible to devote ourselves exclusively to
exhaustive and encyclopedic studies for the next 24 months,
t is highly likely that at the end of that period many of the
problems that we would be asked to help with would have
changed so that our results would no longer be particularly
applicable.
On the other hand, if we succumb completely to the very
pressure upon us to answer all current requests for prompt
:r,'ormation, we will never have any information better than
Ui- slim fragments that we can now supply. Thus our dilemma
it. in a sense, whether to be encyclopedic and irrelevant or
;irrational and incompetent.
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Clearly the only tolerable solution is a compromise betty"
these two extremes. We must try to answer the most im
tant of the problems put to us from day to day as quickly
as competently as possible. But we must reserve a major
of our energies for improving the foundation of knowl
from which better quick answers can be given.
The necessity for this compromise has two further impli
tions. The first is that it is possible to pursue this twof
objective only if we have a certain minimum of resear
A
resources substantially larger than that which the US Gove
ment has allocated to these problems in the past.
The other implication of our compromise is that since
cannot hope to have enough resources fully to exploit all
available information about the USSR, we must be very
that we use our scarce research resources to fill in those ar
of our ignorance which most seriously limit our estima
ability. We must concentrate our scarce manpower on fin
out those things that the US Government needs to know m'
The identification of these priority areas is one of the n~
puzzling problems facing intelligence.
How do we determine basic research priorities?
The most seductive answer to this question is contain
what we may call the "bottleneck fallacy." Since econo,
cold or hot, was first thought of, economists
warfare
,
sought for the bottleneck, the single critical item, the
facility without which the enemy's military economy wo
collapse. The history of the search for such bottlenecks<
record of failure, confirming the economist's faith, that, g
`
a little time resources are highly substitutable one for ano
I
ki
This does not mean that economic warfare is bound t
ineffective, On the contrary, the very fact that resources,
uauu
interchangeable means that to deny an enemy any reso
is to weaken directly or indirectly his military potentiaL>,!
is particularly true in an economy which, like the Soviet
economic activity recognized by the Kremlin as not esig'
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
to these goals would have been abandoned long since. Thus
wherever we make an economic attack upon the USSR, it is
likely to hurt. But it is a delusion to expect that a limited
attack upon a small segment of the Soviet economy will cripple
Soviet strength. It is not the capacity of a particular facility
or the availability of a particular commodity which ultimately
limits the capabilities of the Soviets so much as their total
resources and their ability to organize them effectively.
This does not mean that all things are equally important.
The selection of the more critical commodities and industries
is one way of cutting the problem down to size. But when
one has done all the pruning possible, the number of critical
sectors of the economy remains too great to tackle them all
exhaustively at once.
A second method of determining priorities for research is
to see what basic research would be most relevant to the
problems to which we are being asked to give current answers
now. The dangers in this problem-approach to priorities are
obvious. It leads one always to concentrate one's research
on yesterday's rather than on tomorrow's problems. Basic
research, by definition, takes time. The problems which may
tx~ urgent when the basic research that we start today is
finished cannot be clearly foreseen and are almost certain to
tx. different from those which are plaguing us now.
Furthermore, any attempt to list even the most urgent of
the problems facing us at the moment reveals how many there
zre and how much of the total world economic picture is rele-
vant to their solution. As part of our study of foreign economic
intelligence for the National Security Council, the Central
Intelligence Agency attempted to outline the requirements
for such intelligence in terms of current problems. A very
incomplete sample yielded a list of 42 top priority problems;
u ne of them as broad as t
USSR he totalmilitary potential of the
R.
Again, we cannot wholly discard this criterion We mu
t
s
. "Y to foresee tomorrow's problems and guide our research
accordingly. There are some aspects of the Soviet economy
MORI/HRP THIS
PAGE
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
14 SECRET
which we can take the risk of neglecting. But we must do
mewhat against ti
broad enough job to hedge ourselves so
errors in our own forecasting.
is to take some aspect of the study
lit
y
A third possibi
each commodity and concentrate on that aspect alone for
e could take so
_
commodities across the board. Thus on
ic Commodity Study used
B
as
section of the Outline for a
the preparation of the inventory and fill in that section A
_ __ 4---o ,n,t;l later. One coil
-
in
l
g -- -
ea
every item 111s ,
ies of ORR to the study of require
n
devote the entire energ
io
;
production
for example, or to techniques and methods of
or to levels
dustry
h i
,
n
to the organization and plans for eac
YY
This principle of selection is almost certain to be unsa
ers to most of the questi+
factory by itself, since the answ
which policy makers are going to ask involve putting togeth
- 1W
onc
t at the
a
all of the parts of a basic study to ge
nf the balsa
iti
i
l
an estimate of caps
uirements to achieve whatever
d re
q
between supplies an
lans of the Soviet rulers. An estimate
be the goals and p
vulnerabilities involves a knowledge of the availability
mate ofd
--A -Ion
o
L . ~-
rials at prese~it, pr
goals and plans would be affected if that availability went
int far below requirements. If
be cut by our action to a po
information at
drawn
l
,
y
design of a basic study is proper
ired to arrive at concluq
all the parts of that study is requ
if satisfactory
el
ti
y
r
and no single part can be left out en
ation of each of these methods of dete
ti
g
The inves
h time leads us back to the un<
our researc
priorities on
able conclusion with which we started -namely, tha
f most of the Da
i
'
o
s
encyclopedic and exhaustive analys
in which we can
the only wa
i
s
the whole ecuriom
y
wers to the questions th
sound and authoritative ans
being asked. But we have already determined that we d
g
r
have the time or the resources
end. iHo
to
beg ng
systematic basic studies from g
dilemma?
can we resolve this puzzling
SECRET
MORI/HRP THIS PAGE
The answer is suggested by looking at the present state of
our knowledge. What we have just proved is that we need to
know something about most aspects of most sectors of the
Soviet economy to make a sensible estimate of capabilities,
vulnerabilities, or intentions. But we have not proved that
we must know everything about every aspect. What we already
know permits us to set certain outer limits to the area of the
possible. We know the Soviet Union is at least capable of
certain minimum actions, and we can set certain ceilings on
what they are at most capable of. Our problem is to bring
the "at least" and the "at most" closer and closer together.
This calls for a research program guided by what we may call
the Method of Successive Approximations.
The Method of Successive Approximations
The first step in the Method of Successive Approximations
is to lay out in general terms the specifications of what you
would like to know. What is the list of all the significant
industries, commodities, and services which should be studied,
anti what are the principal problems about them which we
would like to solve? This was the first assignment in our
inventory and resulted in the outlines produced as a guide to it.
The second step is to see how much of the outline you can
fU in and with what degree of precision. This will reveal that
our information about some aspects of each of our problems
better than our information about other aspects. It may
not be very good. The best information that we possess may
rave a very wide margin of error, but other parts of our outline
will be still weaker. Our inventory was designed to bring us
t i ough this second state - to tell us what we know and what
do not know about each of our major problems with respect
to thv USSR. It has revealed what it was intended to show-
4Wneiv. that our ignorance of certain important matters IS
much greater than our ignorance of others.
Tl1e third stage of our Method of Successive Approximations
11 to concentrate our most earnest efforts for a brief period
Ci the important parts of our problem which we know least
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
16 SECRET
This does not mean that we seek authoritative or
t
abou
. answers in these areas of ignorance but merely that we foe
hntil -
on tem u to or somewhat better than our knowledge of the other pa
of the picture.
we have been working in this manner on weak spot
Wh
en
for a period of 2 or 3 or 4 months, we must stand off and talc
ot~
n to the
ti
l
t
~
o
a
another look at where we are then in re
The weak
_
li
e
ne.
"r - --
UUt
else, or we may have gone far enough with them so the
will
the
th
'
y
em,
although we still do not know much about
1-+ a.nrlPnra
s this o
our second over-an look reveal
tackle whatever other sectors of our problem are now tl1
a
ain --
n
g
ot
weaaw~,
out everything about them, but only that we are going to wo
tter re
uir
q
.
on them until our ignorance of some other ma
1;+4
i
ng
know
more pressing attention. In this business,
m
n
a
,7
about a great
knowing everything about a very few things and nothing abo
.,,.a. of iannran
ach substantia
others. +.
tensive enough and substantial enough to pe
i
b
n
e
must
ress toward solutions and not merely
l
prog
make rea
us to hold our own. On the other hand, it must not be pursu+
areas
th
er
with such perfectionist zeal that we neglect o
i
ous.
which our ignorance may be only slightly less ser
In summary, the Method of Successive Approximations
44- ? tarn
-
VU1ves a repeated -/ ---
and several months' production followed by another
in the light both of progress and of changes in the ch
of the problems to be solved.
Problems in applying the Method
In attempting to apply the Method of Successive Appra
whir
i
se
tions, certain common problems and difficulties ar
A particularly bothersome problem is that the things; W
hich it is
i
ngs w
we know least about, and thus the th
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
important to study, are likely to be the things on which we
have least information. In general, we know more about rates
of production of important commodities and products in the
Soviet Union than we do about patterns of distribution of those
products. This is partly because much more evidence is
available on rates of production. The temptation is to study
the material that we have and draw such generalizations from
it as it seems to contain.
In terms of getting answers to our vital problems, however,
we cannot permit the available evidence to dictate the nature
of our inquiry too completely. Several weeks spent searching
for every possible way to button down an illusive fact by
in;enious reasoning from other related facts, by working out
limits on what its magnitude could possibly be from what we
know about other parts of the economy, or by laying on collec-
tion requirements may be worth many times the same amount
time devoted to extracting, setting down, and presenting all
the facts that may happen to be in a given body of documents.
Both methods must be employed. Until we have systemati-
cally examined the available material, we do not know what
can be got out of it. But the material available was not
(:rsigned to answer our questions, and it must be made to be
nm servant of our investigation and not its master.
An irritating feature of the Method of Successive Approxi-
mations is that it may well involve us in going over the same
material several times in search of the answers to a series of
different questions. This repetition is unfortunate and can
br avoided to some extent by investing some time in indexing
and abstracting. If, however, we examine exhaustively all the
material available to us for every implication that it contains
the first time we study it, we will not complete our investigation
?'(3 many, many months. It is unfortunate that research by
the Method of Successive Approximations involves some waste
And some repetition, but it is better than being able to produce
n, answers until 1954.*
t.ditors Note: It is worth reminding the reader that this paper was
I,-,f-pared in 1951.
MORI/HRP THIS PAGE
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
The natural instinct of the researcher who has plenty of tip
is to follow the logical process of trying to build up a pictti
of a whole sector of the economy by first getting an idea of ea
of its smaller component parts. Thus the logical way to es
mate the value of resources used in chemical production is
find out what resources are used in the production of each,
the many different kinds of chemicals. Again this logical
involves breaking each particular chemical into the quantiti
produced in each specific plant. This suggests that the f!
step in answering the over-all question is to try to identify
the physical producing facilities and their capacities and ra
of operation. In many cases, however, a first approximate
to the aggregate figure can be achieved by short cuts whi
avoid the necessity of knowing what in detail it is made up
Thus one can start, for example, with total resources engag
in chemical production in the United States, or in the t
economy of Nazi Germany, as a proportion of total resoun
One can then consider known respects in which the proport
in the USSR must deviate from these examples. Soap is i
in the USSR, and every household does not have its DDT sp]
Such estimates of the whole before you know the parts usuu
have wide margins of error, but when current problems
,
pressing, they are frequently better than nothing at all
Finally, for this Method to be effective, it should ideallj
applied not simply to ORR's schedule of research produc
but to that of the US Government as a whole. Our delinea'
of areas of ignorance should be on a government-wide lx{
and our production to remedy these weaknesses should
planned in collaboration with other agencies so that w
not all concentrate on the same gaps at once. As the coo
integrated with our production so that the Government)
whole may approach more rapidly an adequate understan
of the Soviet economy.
SECRET 19
ECONOMIC INTELLIGENCE
by Paul W. Howerton
EDrrORS NOTE: This article is based on talks
given by Mr. Howerton at the Industrial College
of the Armed Forces in 1955 and 1956. % We are
grateful to the Commandant of the College for
making the transcripts available to us.
T HE first order of business is to identify the problem of
economic intelligence. It is a problem very like that of
the college economics professor in the classic story about
um examination papers that, year after year, ask the same
questions. The punch-line hardly needs repeating: "in eco-
ntnuics. we never change the questions, only the answers."
This is the problem of economic intelligence. We in the
!ntclligence profession have the questions, which remain rea-
:.0uiabiy constant. It is the answers we have to change.
Through a process of refinement, through successive approxi-
mations, we hope to approach the true picture of the outlook
;u the economic sector of the various nations we are called on
t+) study.
Perhaps the best way to treat the subject of economic intelli-
grnce is to borrow the journalistic breakdown into the five
's and the H: "who," "when," "why," "what," "where," and
how-." The first question is what - what is economic intelli-
I:rnce? It is the appraisal of the capability of a nation to
'. ipi ort a war. This is, to be sure, a simplified definition, but
t covers almost every important aspect of the activity.
rile "Why" of Economic Intelligence
The second question is why - why do we prepare economic
Intelligence? We prepare it because we now recognize that
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
20 SECRET
many of the operations undertaken during World War II wer
to
i
th
y
ng
not successful, or were unnecessarily delayed, ow
f sound is
in
lack o
speech made by General George C. Marshall on 9 Septembe,
id
:
1939, just a week after the war started. He sa
The true philosophy of the maximum war effort of a,
supply power as the nation can support. I suspect thi
For some years now she has been devoting ov
outlined
.
50 per cent of the productive effort of her country, inclu
ing men, plants, and materials, for the preparation of wa%
r
and now actual war. So it follows that she is now gea
up to her maximum effort.
that she was not able
however
tant to note
,
,
It is impor
reach this status overnight. It has taken her some fa
'
or five years of intensive effort to develop the raw-ma
acity to support her maximum effort. It is now g
ca
p
erally accepted as a fact that it requires far more time'
than it does to convert civilian manpower into soldiers
This man was Chief of Staff of the US Army. He was alle
be the best informed man in the country on the capabi
t
o
of the potential enemy, and he said that Germany was th"
+?
St
Th
ra
e
in September 1939, geared to its maximum effort.
t
th
a
Bombing Survey conducted after the war indicated
1939 to 1944, in the fields of explosives, tanks, and air
of the German economy was two to three times. The Bri
.1-IL
with a 200-year tradition of intelligence research behind
said at the end of each war-year, "Germany has now i
her peak." And during every successive year, that pe
surpassed.
This, then, illustrates the why of organized economic in
Marshal Montgomery have listed the essentials for na
MORI/HRP THIS PAGE
security, based on their experience. Montgomery, in August
1947, said he believed these essentials were:
1. Strong national character.
2. Great development of scientific and industrial research.
3. Powerful and well-disciplined industrial power.
4. A regular army.
5. Preparedness.
Three of these five essentials are economic intelligence targets:
-great development of scientific and industrial research" (we
have to know what intentions and capabilities may be for the
future), "powerful and well-disciplined industrial power," and
-preparedness" (both of these fall into the category of eco-
nomic intelligence).
Economic intelligence is, in sum, the appraisal of the capa-
bility of a nation to support a war, also an estimate of its
vulnerabilities and of its intentions. Economic intelligence is,
Indeed, probably the best long-range indicator we have of
Intentions. On the vulnerability side, the intelligence com-
u:unity must have, necessarily, a consideration of exploitable
ulnerabilities - a vulnerability is unimportant unless it can
t?? exploited.
"When," "Where," and "Who"
When is economic intelligence produced? It is produced
or both current and future use. The field of economics, broad.
ac it is, requires an intensive study, sector by sector, in any
f i? en country to determine the aggregate of its economic
;-trntial. Furthermore, economic intelligence depreciates at
a constant rate of, roughly, 20 per cent a year. At the end of
} cars' time, a piece of economic intelligence developed from
date Published this year ll w, rtu only 35
yal wui be woiou Oruy a per cellt of its
1' -sent value.
The intelligence community is charged, then, with keeping
`urrcnt on economic developments within the countries under
`tu(ly. These efforts are by no means confined to the study
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
22 SECRET
of the economies of the Soviet-dominated world. It is equal
important that we be equipped to understand the capabilitu
of our friends and alleged friends. There is, for example, u:
_-,
ably no country more closely associated with the US in tl
protection of North America than Canada. There is probabl
no closer working relationship between any two nations in tj
world. And yet we produce intelligence on Canada and o
facilities for war.
Where is this economic intelligence prepared? part of it
prepared in CIA. Part, in the Department of State. The
tary contributes. In National Security Council Directive N1
15, which has been interpreted by the Director of CIA as DC1
15/1, responsibility for economic intelligence research is all
Bated to the various agencies. The Department of State d
.
with broad-gauge economic policy problems. CIA confines
self to the Soviet Bloc and the peripheral areas which
contribute to Soviet capabilities. The military compon
contribute all the military-economic intelligence that is
necessary to the proper understanding of the capabilities
intentions of a potential enemy or friend. The interpretati,
placed on the happenings of the day are contributed
the-board, by all people who are competent to make such
tributions. Consequently, no single organization can
indeed, does) operate in vacuo to produce economic intellige;
It i
s far too important a subject to trust to a single or
tion or a single individual.
Techniques and Methods of Economic Intelligence Produ
The next question is the one I will treat at greatest 1
how is economic intelligence produced?
The chart (Economy of the USSR) * purports to analyze
the economy of the USSR and most of the illustrations 3-
use are drawn from studies of the USSR. The appiicatia
the techniques. and methods discussed, however, is as b:
as the subject matter of economic intelligence. I will
* This chart was devised by my good friend, Bill Tidwell of
Author's Note.
475
?t
MORI/HRP THIS PAGE
METHODS OF
DETERMINING THE ANSWERS
- Statistics
Factory Markings
Sovmat
Input-Output Tables-
Gross National Product
Index of Industrial Production
Population Curve
Index of Standard of Living
SOURCES OF DATA
Attache and Embassy Records
Publications
FBID
Defectors
Returned PW
Commercial Contacts
Travel Folder Program
US Government Files
Covert Sources
Intelligence Services of
Friendly Governments
MORI/HRP THIS
PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78T03194A000100010001-2
the ues-
methods that avea been developed for answering
makers and
telli ence community by policy
is put to the
drawn Upon-
tile sources of information .
The chart, furthermore, breaks down economic analysis
is into
major subquestions. The first is that of qualitative analysis;
the second and third deal with and the
::alysis; organization of the economy,
fourth deals with the ora m th of the economy.
subquestion, with the grow
quantitative anal
number of techniques are available to q u liih d by the
First of all, there are the statithe USSR, stics published
the statistics
untry in question. In the casual cations discussed below,
by and large and with the q
intelligence agencies have carefully d alyzed these
external
, both for their internal consistency he Soviet
can
A;ticsics nce on the operations of the countries within the SThe consistency of components within an aggregate
c
luc. gainst announcements of future chancres in
isily be checked against of individual compo
well as changes
e aggregate itself as
The consistency of data on all levels, including
and the
:. ates of changes in the physical productive capacity
outputs within the Soviet Bloc, has been verified
R
~n;modity
or a sufficient number of cases to convince us that the ~-
in eneral, to distort their published
sians are not attempting, 9
statistics. examples of precisely the
Having said this, let me give a few opposite - cases where the Russians
understand proper yed the
where interpretation is necessary
statistical analyses made by the Soviet Central Statistical Bu-
reau. A Soviet rubber-produing plant turns out both rubber
tires and rubber heels. An announcement came out of this
plant saying that the goal for tires had been missed by 50 per
cent but that the goal for heels hsda es meant dthat t e over-
cent - which, according to the loo per cent!
all production goal was met by Krokodil.
Another example comes from the humor magazine,
A cartoon appeared in this journal, some time back, which
ger of a machine tractor station standing on
showed the manager
MORI/HRP THIS
PAGE
Approved For Release 2007/10/23: CIA-RDP78T03194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
24 SECRET
the porch sending off a deputy to the local commissar with the
admonition: "Don't forget to tell him that half the plan is 100
per cent fulfilled."
All of which demonstrates that Soviet statistics need care.
ful analysis. They cannot be accepted completely on face
value. But by such analysis, and by careful review of the
aggregates that we are able from time to time to accumulate,
we have come to the conclusion that the statistics are, by and
large, valid.
Input Analysis
A second tool of quantitative analysis is the method of in-
puts. There are certain basic relations, that is to say, that
are constant throughout a given industry. For example, the
floor space in an aircraft plant is proportional to the number
of airplanes it can produce. This relationship has been checked
out in a number of aircraft companies in the US and in friendly
countries and found to be valid. Other methods have been
developed to relate seemingly unrelated commodities - com-
modities which seem to have a mutual control over one another.
An example is steel and rubber. The Joint Intelligence Bureau
in London developed this factor: the amount of steel produced
in a given country is directly proportional to the amount
of rubber consumed. This factor has been tested in the US,
France, and Italy and found to be valid. These are w exam-
ples of how factors can be developed so that, given a bit information collected by an observer in the USSR on some
plant or industry, one can by deduction determine the approxi-
mate production capacity of that plant or industry.
Markings Analysis
In intelligence, as in laboratory research generally, we
try to verify our results by using a number of different
analytic methods. One of the most useful of these methods
is that of factory markings analysis - next item on the list
of tools of quantitative analysis. Factory markings are those
trade marks, inspectors' marks, or other stampings, that appear
on most any finished product. These markings are unique
a given plant or even for a given department or inspector
for within a plant. Factory markings analysis might be called
the Bertillon system for the identification of products and
producers.
Even though the field is in its infancy and was only devel-
oped during World War II, some examples drawn from wartime
experience demonstrate its enormous possibilities. The ver-
age monthly production of tanks in Germany in the period
th-
hadintelligence
140-1942 was estimated, using conventional
analyzed ethe
ods, at 1550. The markings people,
markings on captured tanks, set the figure at 327. When the
Speer Ministry files were captured, the true figure was found
to be 342 - 327 (markings analysis) against 342 (true figure)
against 1550 (conventional analysis). During 1942 the con-
ventional estimate of German truck production was 2 8000,0 00;.
the markings estimate, 97,000; the Speer Mi Ministry figure,
tires was
The 1943 average German monthly production markings
estimated by conventional methods at 1,250,000; by
analysis, 175,000; the actual figure, 186,000. It is, of course,
jlist as bad to overestimate production as it is to underestimate
it. It has been said that the bof
Europe was intelligence and,
unnecessarily delayed a year
especially, bad economic intelligence.
Modern machine methods handle the raw data of markings
analysis in the Joint Markings Center maintained by CIA in
collaboration with the Army, Navy, and Air Force. Markings
analysis is going to be one of the potent techniques of economic
intelligence in the future. Its method is that of statistical
analysis; and, consequently, the larger the sample, the more
accurate will be the conclusion. With five MIG's available to
US intelligence, markings analysis has produced an estimated
monthly production to within one plane of the figure arrived
at through other intelligence methods. Just to illustrate the
dimensions of the markings effort, by the way, there are 36,000
distinctive markings on a MIG, of which 3,600 are significant
for markings analysis.
MORI/HRP THIS PAGE
MORI/HRP THIS PAGE
Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
 Approved For Release 2007/10/23: CIA-RDP78TO3194A000100010001-2
Input-Output Analysis
A system of input-output analysis - or, inter-sectoral anah;
ysis - is useful for quantitative estimation, because it sho*g
the changes that can occur within an economy for any given
stimulus. Take, for example, an input-output matrix for the
Soviet petroleum industry (plotting products on the vertical,
side and industry on the horizontal) : the agricultural sector
within the USSR uses 23 per cent of total petroleum output;
energy production uses 12 per cent; manufacturing, 11 per
cent; transportation, 24 per cent; household uses, nine per
cent; and the military, eight per cent. This adds up to 87
per cent. The other 13 per cent is that bit of petroleum the
Russians have been using for barter with the West for scarce
and necessary machine tools.
input-output on electric power is another useful illustration
of this technique of analysis. In their current plan, the Rus-
sians hope to have 170 billion kilowatt hours of electricity
production a year, which would require 83 million tons of coal
to produce. For each kwh, 1.2 pounds of coal are required.
If the USSR decides it is going to have, say, more aluminum
and will therefore have to increase its power requirement to
200 billion kwh, 95 million tons of coal will also be required.
This new coal requirement of 12 million tons must now
come from new mining activity or must be reallocated within
the present consumption pattern.
As one further example of input-output analysis, consider a
changeover in a given Soviet oblast from horse-drawn agricul-
tural equipment to tractors. This sounds, at first, like a simple
transformation; but, to increase requirements for tractors
means much more than just an increase in tractor production;
it means, as well, an increase in steel production, in electric
power production, and in electronic control; and it means an
increased demand for management and skilled labor. The
reduced requirement for horses, on the other hand, will mean
among other things a larger food supply for the people. It is
just this sort of complex economic interrelationship that input-
output analysis can help to clarify.
MORI/HRP THIS PAGE
In sum, the development of a matrix of input-output will