THE TRADECRAFT OF ANALYSIS CHALLENGE AND CHANGE IN CIA'S DIRECTORATE OF INTELLIGENCE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP98-00412R000300010016-9
• Release Decision: RIPPUB
• Original Classification: K
• Document Page Count: 17
• Document Creation Date: December 22, 2016
• Document Release Date: January 20, 2012
• Sequence Number: 16
• Publication Date: August 1, 1994
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
1 August 1994
Challenge and Change
in CIA's Directorate of Intelligence
Douglas J. MacEachin
Tying the need for change in CIA's Directorate of
Intelligence (DI) to the ending of the Cold War has tended
to cause the focus for change to be on size and subject
matter. There is the not unreasonable belief--inside the
Agency as well as outside--that whatever organization
emerges from the change process must be smaller and that the
needed reductions are to be achieved mainly by cutting
resources committed to the threats formerly posed by the
Soviet Union.
The need for change goes beyond size and subject
matter, however; dramatic changes in national priorities and
the resulting funding reductions require a reexamination of
the fundamental purposes and process of intelligence
analysis. We need to go back to the basic questions of what
we do and what is the best way to do it.
The purpose of this monograph is to describe corporate
efforts underway in the DI to deal with these questions. It
is not intended to be a critique of past practices or of the
quality of analysis carried out in the Directorate. There
has been much more that has been good than we are given
credit for, but there also is much that needs changing.
We have all been a part of what has been good and of
what needs changing. Some of the practices needing change
came into being as the unforeseen result of efforts
undertaken for sound constructive reasons, judged by
participants at the time to have been fully valid, and
usually aimed at correcting deficiencies of the time. But
collectively, with time and bureaucratic evolution, they
turned out to be counterproductive.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
We cannot chart a course for the future without facing
up to existing problems. Designing change requires us to
lay bare what it is about the present system that needs
changing, what problems the change must correct, how the
changes are to make us better. Change in any large
organization comes with substantial cost, and we need to be
able to demonstrate--both to the work force and the outside
critics--that the gains outweigh the costs.
The following sections describe changes that have been
made or are underway in the DI in the fundamental areas of:
-- What we do: Making the needs of the user--the
policymaker--the driving factor in intelligence
production,. defining the contribution to those
needs as the measure of performance, and
eliminating practices that encourage production
aimed at fulfilling internally established
conventions for achievement and recognition.
-- How we do it: Redefining our analytic tradecraft
to emphasize "facts" and the "findings" derived
from them. This does not mean opinions are no
longer valued, but it does require that their,
credibility be established through intelligence
practices that clearly identify what is known,
how it is known and with what level of
reliability; what is not known that could have
importance consequences; the "drivers" or
"linchpins" that are likely to govern the outcome
of dynamic situations; the analytic calculus
underlying all conclusions and forecasts; and the
uncertainties in any of the components of the
analysis and the implications of those
uncertainties for alternative outcomes.
Adherence to these principles of analytic
tradecraft are to be the standard of professional
excellence--the "professional ethic"--of the
Directorate.
These changes are not directly tied to the substantive
dynamic generated by the end of the Cold War. There is in
fact consensus among the corporate leadership of the DI that
these changes have been needed for some time. They
represent major cultural change, however, which entails
sizeable cost and pain, and it has taken the events
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
generated by the end of the Cold War to make clear that not
making these changes ultimately would entail even higher
costs than making them.
-- In sum, the need for change has existed for some
time, but the end of the Cold War clearly
generated force for change and provided the
occasion and conditions for implementing change.
This is not meant as an argument against proposals for
reducing the amount of resources devoted to intelligence and
realigning the substantive focus of those resources for the
post-Cold War world. A major effort is underway to design
the architecture for a smaller DI with more streamlined
supervisory structure and greater organizational flexibility
to respond to more diverse and rapidly changing
requirements.
But without the kinds of systemic changes described
below, we would end up just doing the same things in.
different boxes, and that is not sustainable in a post-Cold
War world. Therefore, while examining issues of resource
allocation and ways to improve efficiency at lower levels,
DI management during the past year has given first priority
to fundamental changes in how we define our mission and
carry out our work.
THE MISSION: FOCUSING ON THE USER
Crafting mission statements is always a difficult and
somewhat metaphysical task. For purposes of this review, a
short statement that is more descriptive than definitional
seems most useful:
Provide U.S. policymakers with information and analysis
they need to carry out their mission of formulating and
implementing U.S. national security policy.
There is not enough space here to try to take on all
the excursions that could be made on this statement. It
leaves open such questions as what constitutes "national
security policy," and therefore which substantive areas are
addressed and which policymakers are the intelligence users.
It also is broad enough to encompass what is often referred
to in intelligence jargon as short-term "tactical
intelligence" and longer term "strategic estimates," and is
also neutral as regards the "clandestine" versus "open
source" debates. Such discussions have bearing on issues
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
such as breadth of coverage, and ultimately on resource
levels, but they do not affect the basic principle of the
statement--that all considerations of the intelligence
mission start with the user, who is the U.S. policymaker.
Some will doubtless argue that there is nothing new in
this, that this has always been our understanding of our
mission. Indeed, if a decade ago the question had been
posed to a random sample of DI officers, the dominant answer
probably would have been articulated much along the lines of
the statement above. A fundamental tenet of my own argument
for change, however, is that back then the statement did not
always drive what we did in practice nor how we did it.
Although great strides were made in the 1980s toward
exposing our officers to the policy world and gaining
greater understanding of its operation and needs, our
business practice for a time evolved in ways that allowed
our own internal goals and measures of quality to have a
greater influence on our product than did the questions of
"who is my user" and "what does my user need." In recent
years, we have pushed the pendulum back toward direct
customer service and we are now accelerating that trend.
The Publish or Perish Factor
There are probably some who will want to contest this
criticism, but most present and past officers with whom I
have tested it have agreed with it. A simple description of
the dynamic that existed makes its own best case for the
charge:
-- Success in the DI meant showing a "production
record" that included published papers. But to
get their products published, analysts had to get
them through a review process that contained at
least four and-sometimes more layers, from the
immediate supervisor to the head of the
Directorate. Not every written product had to go
through this gauntlet, but those that did not
counted for less in the rewards and advancement
system.
-- Once a paper cleared this process, it was
disseminated according to a list of "user
profiles" maintained in the office that did the
formal printing. Publication was, de facto, a
decision taken separately from a detailed
identification of the audience.1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
The published products that counted most for
demonstrating analyst achievement--and which therefore
counted most for analyst advancement--followed the general
rule that long was better than short (despite protestations
to the contrary) because it demonstrated full scholarly
effort. The Intelligence Assessment (IA) which was both
long (incorporating extensive research and description) and
"analytic" carried the most prestige within the DI culture.
It was considered more reward worthy than the Research Paper
which, while as long or longer, was more informational and
descriptive than judgmental. Both, however, outranked--for
purposes of rewards and promotions--the shorter memoranda,
despite the fact that our most important consumers have
consistently told us that it is the latter that they found
most useful.
Production criteria for advancement also included the
need to have produced a certain amount for the daily current
intelligence publications. In some substantive areas, a
high volume of current intelligence production could carry
an analyst a certain distance up the promotion scale, but
sooner or later the analysts had to prove their fundamental
mettle by publishing one of the "big" papers.
-- Analysts knew that to advance their careers they
had to get papers published, and they knew what
kinds of papers counted most for this purpose;
managers knew they had to demonstrate component
production records.
Most of this was instituted for what were seen at the
time as sound, constructive reasons. For example, the IA
was indeed the most challenging product of the DI in terms
of demands placed on the individual producer, but it
reflected a "degree of difficulty" system of evaluating
products, rather than utility to the consumer. The layered
review system was in large measure a version of the very
appropriate principle of management accountability, and an
effort to impose more rigor on the products.
-- The way that these factors came together should
make all of us who were here through it all find
some sympathy with the Alec Guiness character in
The Bridge Over the River Kwai.
-- Our only solace lies in surveys of other
organizations in the private sector, which
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
disclose that the tendency to drift away from
customer focus toward internal criteria is one of
the most common problems of large organizations.
Changing the System
In the mid-to-late 1980s, many efforts were made to
address these problems and there was particular progress
made in diversifying the product line. This trend was
further accelerated by the Production Task Force, one of
many task forces set up by Director Gates during 1992.
An important step was taken by John Helgerson at the
beginning of 1993, when he eliminated the mandatory seventh
floor review of finished products. Since then, the
individual offices have taken responsibility for the review,
and they have instituted further flexibility on the level
and amount of review required. The principle of management
accountability still demands review of most products at some
level, but the effort is to have it done.as low in the
supervisory line as possible, commensurate with the ne5d for
timeliness and the sensitivity of the issue addressed.
-- Today, the DI front office engages in pre-
dissemination review only on request--if, for
example, a production component wants to get
another view on a particularly sensitive or
venturesome product--or in the event the product
is one that was directly tasked by the DDI or
ADDI.
-- Over the past year, these factors have led to DI
review of no more than about a dozen or so
products.
The practice of drawing up DI publication target lists
for each fiscal year has been discontinued. Any publication
lists that are prepared are done at the production office
level, and they are used only for purposes of office-level
planning and for coordinating resource use with other
intelligence producers such as the Defense Intelligence
Agency. No percentage tallies are kept; the focus is on
quality and impact, not numbers or quotas.
The annual Directorate-level planning exercise which
formerly produced the publication targets now focuses on
strategic lines of analysis as a basis for resource
allocation decisions and collection and research planning.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
Research strategies developed in this effort are designed as
capital investment in our intelligence production
capabilities, not as a publication rationale.
Some in the DI have expressed a concern that the
changes mean less value placed on research. That is an
incorrect perception. What is being done is to establish a
distinction between "research" and "publication." Research
is critical to our effectiveness, and constitutes capital
investment. What we are is what we know, and how well we
analyze what we know. But the standard we want to apply to
assessing the value attached to any given research effort is
the knowledge gained and its contribution to our mission as
a whole, not whether it fulfills a publication target.
-- The results of individual research efforts
probably will as a rule be compiled in some
written form to exploit the learning and
knowledge sharing benefits, but dissemination in
hard cover form will be a separate decision based
on demonstrated need by the policymakers.
-- The results from a single research project can be
disseminated in one or more of several forms:, a
relatively long written compilation of the
information for other members of the Intelligence
Community and selected components of the policy
community that engage in in-depth studies; a
short Intelligence Memorandum, a few pages in
length, describing for more senior policymakers
the principal findings and implications of the
larger body of research; and a one-page version
for the Cabinet Principals.
-- Other research efforts might result in one or
none of these products, and instead become part
of a data base.
The DI will continue to devote resources to some
products that are disseminated to a broad and diverse
audience, including the NID and various specialized periodic
journals which have high value to the users. We also will
do our long-term strategic assessments when there is a clear
need. The key to all of these, however, is demonstrated
need by the user, not continuation of a routine activity on
our part.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
The most important strategic change has been directed
at institutionalizing the principle that our analysis is for
people not publications. The measure of achievement is not
the publication of a paper but the delivery of intelligence
information and analysis to a user that is critical to the
users' needs. Written products are a means and not the only
means of communication; face-to-fa5e briefings are in many
cases more useful to the consumer.
This means customizing products according to issue and
user. A paper or briefing that is designed to be useful to
100-200 people is not likely to be of significant value to
the people--ranging from a handful to perhaps 30 or so--who
are most deeply engaged on a given policy issue.
These officials are almost always going to be up-to-
date on events that affect their most pressing policy
concerns. They have staffs dedicated to keeping them
informed, and they have ready access to reporting through
diplomatic and defense channels, open sources, and even raw
intelligence. They also routinely talk to foreign
counterparts, journalists, academicians, contractors, and
lobbyists--to name a few of our competitors.
Intelligence products for such consumers must be
governed by the principle of "value added." Accumulation of
information and views is not sufficient by itself to warrant
a publication. Unlike the publish or perish system, more is
not a criterion for better. The threshold for production is
defined by the needs of a specific set of users, and the
contribution of information and analysis to those needs--not
whether the product qualifies for the publication
scoreboard.
Moreover, the written product on a given issue need not
and often should not be the same for each user. For
example, an intelligence report or memorandum prepared for
the Deputies and IWG members would properly be paralleled by
a one-page highlight version for Cabinet Principals.
FACTS AND FINDINGS: THE PROFESSIONAL ETHIC OF ANALYTIC
"TRADECRAFT"
For a long time it has been an article of faith in much
of the DI that the primary value added from our product--
written or oral--is our judgments. Consumers present and
past, however, have been virtually unanimous in emphasizing
that the first element of critical importance to them is
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
that we give the facts. First and foremost, they want the
answers to the questions what do we know? How do we know
it--is it from a spy, an intercepted telephone message, a
picture, or an embassy report? What important factors don't
we know? The products also must make clear the critical
difference between what is, versus what a source reports, or
national technical means indicate, or diplomatic observers
calculate.
The reaction of some DI analysts to this is to question
whether it means that their opinions are being devalued. On
the contrary, opinions are indeed highly valued, but to be
so they must be anchored in sound and clear argumentation
that lays out what is known and with what degree of
reliability, what is not known that could have significant
consequences, and the analytic logic that forms the basis
for the findings.
Findings are critically dependent on the analysts'
ability to bring their expertise and knowledge base to the
evidence. The patterns, relationships, and trends that
analysts detect in the evidence are combined with the
knowledge they bring to bear from both their training and
education and from their daily scrutiny of the endless flow
of classified and open source material. And this process
needs to be explained to the recipient of the findings.
-- "How do things usually work" in country "X"
regarding weapons acquisition or economic
corruption or manipulation of elections?
-- "What is increasing, decreasing, new, and
different" regarding terrorist incidents,
diversion of resources to military objectives,
money laundering?
-- What is the relationship between meetings,
speeches and action on the part of leader "Y"?
Some have suggested that including descriptions of the
evidence and sources of it and laying out the analytic
calculus is contrary to the need to make the products short
and sharply focused. In fact, our experience in
implementing these principles has shown that this is not so.
A snappy march through the evidence followed by a
delineation of the analytic logic for any conclusions makes
for a short, punchy paper. What has tended to make papers
long has been the practice in the past of writing them for
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
general audiences, thus requiring long background and
explanatory sections that the informed, engaged policymaker
does not need on his or her issue. (These background and
educational contents were also needed for the several layers
of review that had to be cleared.)
Some DI analysts also have expressed concern that
laying all of this out in their products will lead the
consumers to do their own analysis. Most consumers,
particularly those directly engaged on a policy issue, are
already doing that with or without the intelligence product.
They have their own channels of information, often they have
much of the same raw intelligence possessed by the analyst,
and they often know the main foreign players from direct
interaction.
Consumers present and past have consistently told us
that for them, the value added--and the credibility--of the
intelligence product is directly dependent on the
information conveyed, its reliability, and their
understanding of the analytic logic that supports the
conclusions. If these are not made explicit and clear, the
intelligence product becomes simply an opinion that may be
agreed with or swept aside.
Forecasting vs. Fortunetelling: Identifying the
"Linchpins"
Most consumers understand the difficulties of dealing
with the unknowable. The future cannot be "known," nor can
we "know" what is in the mind of leaders. These things can
only be forecast or estimated on the basis of the evidence
available and the combined impact of knowledge and
expertise. But laying out the evidence and showing the
interrelationships that form the basis for the judgments can
define the difference between predictions derived through
forecasting and predictions derived through fortunetelling.
This distinction is not a matter of forecasts
always providing a correct answer and
fortunetelling an incorrect answer. With
unknowables, there will always be forecasts that
turn out to be incorrect.
The difference is that in forecasts the
evidentiary base and analytic logic is
transparent.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
Analyses of potential developments are based on
assessments of factors which together would logically bring
about a certain future. These factors are the "drivers" or
"linchpins" of the analysis. If one or more of them should
change, or be removed, or turn out to have been wrong to
start with, the basis for the forecast would no longer hold.
-- Identifying the role of these factors in the
analytic calculus is a fundamental requirement of
sound intelligence forecasts. The policymaker
needs to know the potential impact of changes in
these "linchpins."
-- The consumer especially needs to know if there
are any of these "linchpins" for which the
evidence is particularly thin, for which there is
high uncertainty, or for which there is no
empirical evidence but only assumptions based on
past practice or what appear to be logical
extensions of what is known.
A review of the record of famous wrong forecasts nearly
always reveals at least one "linchpin" that did not hold up:
The Soviets will not invade Czechoslovakia because they will
not want to pay the political costs, especially after having
signed the Rejkavik Declaration the previous year; the
Soviets will not invade Afghanistan because they do not want
to sink SALT-II, which at that moment is being debated by
the U.S. Senate; Saddam Hussein needs about two years to
refurbish his military forces after the debilitating war
with Iran and therefore will not, despite evidence of
motives for doing so, invade Kuwait in the foreseeable
future.
-- Each of these linchpins was more of an assumption
or postulate than an empirically based premise.
Each was logical, but each turned out to be
wrong.
-- In each case, the sin was less in the fact that
the linchpins did not hold than in the failure of
the intelligence products to highlight the extent
to which they were assumptions, and the potential
impact on the bottom line judgments if they did
not hold.
This approach to analytic tradecraft--facts and
findings, and linchpins--forces systemic attention on the
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
range of and relationships among the factors at play in a
given situation. Laying out the linchpins encourages
testing of the key subordinate judgments that hold
estimative conclusions together. Since the premises that
warrant the bottom-line conclusion are subject to debate as
well as error, analysts marshal findings and reasoning in
defense of the linchpins. Finally, this approach helps to
focus ongoing collection and analysis: what indicators or
patterns of development could emerge to signal that the
linchpins were unreliable? What triggers or dramatic
internal and external events could reverse the expected
momentum?
Implementing Change in "Tradecraft"
For most of the past year the application of these
analytic tradecraft principles has been the primary focus of
DI supervisors. This has included redesigning the product
formats themselves in ways that facilitate customizing for
users and create a natural flow for the "facts, findings,
linchpins" tools. The DI's own Product Evaluation Staff has
disseminated guidelines that establish these principles as
the standards against which products will be evaluated, and
as the measure of excellence for evaluating analysts',
professional performance. All of us believe that much
progress has been achieved, that the principles have begun
to take root in the Directorate, and that change is evident.
Our aim, however, is to ensure that this is not allowed
to be a passing thing, but rather becomes embedded in the
professional ethic of the Directorate of Intelligence. To
this end, a "Process Action Team" was commissioned in the
latter part of 1993 to study means of institutionalizing
these "analytic tradecraft" values in the Directorate
culture, with particular emphasis on training supervisors
and analysts. The Team was made up of supervisors and
analysts from various substantive areas and echelons, and
undertook an intensive training session with outside
consultants on quality tools that addressed problem solving
techniques.
This Team's efforts included the canvassing of
nearly 200 intelligence officers to understand
their perceptions and expectations on the
definition of DI tradecraft and the ways in which
it had lasting impact. In these discussions the
Team gave particular focus to officers who had
served in assignments at policy.agencies, where
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
they had the opportunity to see firsthand the
needs and interaction with intelligence.
-- The Team also undertook extensive discussions
with people from other government agencies,
private industry, the media, law and medical
schools, and philosophy departments at colleges
and universities to learn the systems that had
been most effective.
The final report from this Team has been disseminated
to all DI Officers and has been followed up with extensive
oral briefing to DI supervisors and analysts at all levels
as well as to nearly all senior Agency managers. It is in
the process of being presented to the Staffs of the
Congressional oversight committees.
As a result of the findings of this Team effort, the DI
has cancelled all runnings of the current curriculum of its
analyst training courses. A DI Training Board has been
established to oversee the complete reengineering of the DI
officer training program, for use at all levels including
the senior-most supervisors.
The new curriculum, scheduled to be ready to launch
around the end of the year, is being designed by DI officers
detailed to this assignment, and reviewed by the DI
corporate management with outside consultants. The courses
will be taught by DI managers and analysts, as well as some
training specialists, on the theory that teaching is one of
the best ways to embed learning. This also will be aimed at
ensuring there is no fault line between the training and its
application in the production unit. The courses will use
real world cases drawn from past and present issues and
products.
-- Studying what is good or bad about products from
the present work force on issues directly
confronted by our current consumers should
provide extra stimulus for the application of the
principles in the day-to-day work of the
Directorate.
-- Training has the most impact and lasting value
when it is part of the common, everyday
professional activity of the work force at all
levels, when they engage in it together as
colleagues.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
Finally, rewards and recognition other than promotion
must also be brought into line with the professional
standards and values we wish to encourage among our work
force. A Process Action Team is concluding its study on how
we can bring this about and will soon make its
recommendations.
Attitudinal or cultural change is not possible unless
the behaviors we seek are openly recognized and rewarded.
In an age of information explosion, bringing value added to
our customers will require a blending of skills. Analysts,
data organization specialists, visual display experts,
research assistants, to name a few, will no longer be able
to work independently, but must be organized into a skills
team. Synergy and efficiencies in the work process are a
must if we are to meet demands with a sharply reduced
workforce.
A WORD ON "POLITICIZATION"
The close engagement of the policymaker-consumers and
the more customized products sought through the changes
described above have raised for some a concern about the
risk of politicization. Fundamentally, the only guard
against politicization'lies in the integrity of the
supervisors and analysts themselves. Formulating the
products for general readership and distancing of analysts
and firstline supervisors from policymakers is no guard
against judgments being skewed to fit someone's particular
vision.
The tools of tradecraft described above, however, do
provide a standard which is inherently resistant to skewing.
Conclusions are to be presented as the result of evidence
and analysis, not simply as "views." Disagreements must be
focused on the evidence and logic, not the judgment that
proceeds from them. Conclusions stand on how they are
derived, not simply on who made them. If a reader or
reviewer who dislikes or disagrees with a given resultant
conclusion is to be credible as offering more than a
parochial challenge, he or she must be forced to show where
the facts or the analytic calculus is wrong or how their own
analytic calculus is the correct one.
-- When analytic judgments are fully and
demonstrably backed by evidence and the logic is
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
explained, simple assertions of disagreement
don't do much for the asserter.
-- On the other hand, conclusions not backed up by
the analytic tradecraft become of the nature of
assertions themselves, and invite the counter-
assertion from whoever sees the conclusion as
unhelpful.
In a large proportion of cases, the sketchy evidence
available can in fact support different conclusions,
depending--for example--on the unknowns and the "linchpins."
This is what alternative judgments are all about. The
intelligence analysts themselves should--indeed must--be the
people to show the alternative outcomes and what are the
factors that drive the-differences. That is a fundamental
principle of good intelligence Tradecraft; it is what
linchpins are all about.
Experience has forcefully demonstrated that however
much policymakers may differ on political philosophy, they
react much the same to intelligence reports that run
contrary to their expectations or hopes. Unhappy responses
come with the intelligence territory. No policy official
likes to see intelligence that suggests things are going bad
for the policy, any more than a coach likes to hear a
prediction that he will lose the game. But the job of a
football scout is not to turn in a prediction of the score,
but to give the coach information useful in formulating a
game plan. The same applies to intelligence analysts.
Objectivity is not a matter of who makes the judgment;
analysis is not objective simply because it originates from
an intelligence analyst, anymore than it is automatically
nonobjective if it originates from a policy official. Bias
is not the private preserve of anyone. Analysts can become
wedded to a certain analytic view, just as policymakers can
become wedded to a policy view. The principles of
tradecraft are the means that help all of us avoid falling
victim to our own perspectives.
The
test
of objectivity is how well the evidence is
laid out
and
its reliability exposed and how openly and
carefully
the
basis for any conclusions is presented.
Analysts
who
are confident that their information is fully
laid out
and
their analytic calculus fully presented have
nothing to fear from someone's attempt to find other
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
conclusions. We can let the quality and objectivity of our
work speak for itself.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9
NOTES:
1One former CIA officer has noted that this system on more
than one occasion resulted in a product being disseminated
to a person who was deceased or otherwise departed.
2One of the specified objectives in the studies of potential
changes in organization currently being conducted is to
facilitate the "flattening" of the review process while
maintaining the benefits of management accountability.
3During March 1994, the DI delivered briefings to policy
officials at the Deputy Assistant Secretary level or above;
552 were on demand.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/20: CIA-RDP98-00412R000300010016-9