MEMORANDUM FOR: SENIOR REVIEW PANEL MEMBERS FROM DIRECTOR OF CENTRAL INTELLIGENCE

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP86B00269R001100100007-1
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 22
• Document Creation Date: December 19, 2016
• Document Release Date: January 10, 2006
• Sequence Number: 7
• Publication Date: December 19, 1983
• Content Type: MF

SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
ER 83 0093/1
Study - Notes1
cn Cuba & Nicaragua
MEMORANDUM FOR: Senior Review Panel Members
FROM: Director of Central Intelligence
Following on my earlier memorandum today, this analy is of work on Cuba
and on Nicaragua in 1979 and 1980 may help you address those items in the
context of this memorandum.
William J. Casey
Attachment:
Memo to DCI from NIO/LA, dtd 15 Jan 1982,
Subject: Review of 1981 Cuban Estimate
and Past CIA Analysis on Nicaragua--Summary
of Comments and Suggestions
Approved For Release 2006/01/3Y%kRDP86B00269 R001100100007-1
 Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
15 'January 1982
MEMORANDUM FOR: Director of Central Intelligence
FROM : National Intelligence Officer for Latin America
SUBJECT : Review of the 1981 Cuban Estimate and Past CIA Analysis on
Nicaragua -- Summary of Comments and Suggestions
1. Responding to your request, two personal reviews of some past CIA work
on Cuba and Nicaragua are attached:
Toward Latin America, including an
- Cuban Policy
13 Jan 82
Tab A
,
Appendix (pp 9-13) summarizing the changes in the
11 drafts of the NIE from Jul 80 to Jun/Jul 81.
including Appendix A (pp 5-8)
uan'Review
Ni
15 Jan 82
Tab B
,
-
carag
summarizing all the formal production items and
Appendix B (pp 9-12 ) summarizing the informal
memos from Jan 79 to Jul 81.
This draft has not been reviewed by or coordinated with any office at this stage.
2. For each draft, the following brief overview will summarize my comments
and several related suggestions for "how we might do better in the future".
Comments on the NIE: Cuban Policy Toward Latin America ,at Tab A)
COMMENTS ON SUBSTANCE
Estimate dr.a.fts were behind events and data
-- e.g., initial Jul 80 draft far more
tentative than CI A data and 2 May 79
analysis;
-- e.g., final Jun/Jul 81 estimates did
not take account of events in spring 81
or the massive synthesis of data on
Cuban covert action done by the intelli-
gence community in May 81.
- A "mind-set" problem _
-- e.g., drafts saw domestic failures,
desire for "accommodation with US",
policy "in transition" factions first
preventing then impelling Cuban
aggression .
SUGGESTIONS
-- After terms of reference are agreed,
but before estimate drafts are
written, prepare factual
summaries, chronologies, maps,
lists of previous relevant CIA
publications.
-- Establish clear responsibility for
updating the facts on a monthly
basis.
-- Drafts should clearly summarize
the factual information on key
questions
-- Shen provide for alternative per-
spectives, even if in appendix
only.
CL BY SIGNER
Approved For Release 200?g Ml
:CIA-RDP86B0 269R00110 b,0 M~OL02 PLE 25X1
I FM
 Approved For Release 2So6994
COMMENTS ON SUBSTANCE
Cuba-USSR relationship initially over-
complicated
e.g., initial focus on doctrinal
differences while ignoring the DGI/KGB
relationship, unity of objectives, etc.
- Parochial failure to take account of Cuban
actions in Africa Middle East and their
implications
SUGGESTIONS
-- Where CIA has done a great deal
of previous work on a key issue,
the draft should briefly
summarize those conclusions
and then agree or disagree
with reasons--why the specific
situation is different.
-- Obvious need to use previous CIA
analysis to explore what new
skills, methods Cuban forces
might have learned in their
extensive Africa/Middle East
,cticins.
Comments on Nicaraguan Analysis, 1979 to July 198at Tab ~J
e.g., discussions focused on Cuban
actions in Latin America virtually
ignoring its massive activities in
Africa and the Middle East for their
implications about Cuban purposes and
techniques.
COMMENTS ON SUBSTANCE
Good performance in:
-- warning of the impending crisis;
-- monitoring extensive Cuban support to
and influence on FSLN;
-- immediate post-victory Cuban/Nicaraguan
decision to help the Salvadoran
revol utionary forces.
Medium performance:
-- too slow on likelihood of FSLN victory;
-- too reluctant to draw the implications
of Sandinista moves to consolidate
power;
SECRETA
SUGGESTIONS
-- Write an alternative, more
pessimistic analysis with
indicators.
-- Analysts working on a new Marxist
revolution should look back at
Cuba and other cases of power
consolidation (Poland, 45-48)
for perspective.
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 Approved For Release ?&&BfO 30 : CIA-RDP86 0269R001100100007-1 25X1
COMMENTS ON SUBSTANCE
-- unwilling to assess the implications
of the immediate Cuban/Nicaraguan
cooperation with El Salvador
guerrillas.
Key omissions
-- much too little reporting and analysis
of the Mexican/Socialist International
support for revolutionary left;
-- too little attention to the Palestinian
terrorist/radical Arab connection;
-- much too little information on internal
and international moderate groups and
countries (has led to repeated CIA
error "the Europeans are all against us
in El Salvador"--failing to distinguish
between the Social Democratic and
Christian Democratic positions).
SUGGESTIONS
-- Analysts should see this phase
as a struggle (without guns)..
ano report on the assets and
resources of both sides and
look ahead.
Role-lay the enemy--sketch what
might be done based on evolving
patterrs.
more realistic and wholistic
understanding of political
and paramilitary war is clearly
needed--the role and importance
of international propaganda,
meetings, deception, front
groups--some past case studies
should be synthesized (Algeria,
Greece, Philippines, Vietnam)
for broad lessons.
Atts: Tabs, A & B
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRETA Tab A 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
3 January 1981
Cuban Policy Toward Latin America NIE 8 J/0 9-81I
Comments on the Drafts from July 1980 to July 1981 *
During a one-year pcriod,starting in July 1980, 11 successive drafts of
a National Intelligence Estimate or, C-:ban Policy Toward Latin America were
written and reviewed within CIA and the community. Fina' products were issed
in June and July 1961. The Director has asked me to review these drafts and
"analyze what we can learn in order to do better in the future".
In answer to the question "why the estimates. . . were altered as
successive drafts were created. . . .", the evidence from the texts is that
various reviewers (including the then NIO for Latin America) were not in
agreement with the views of the drafter nor, as it turned out, was the initial
author who made substantial changes in his judgments in the face of Cuban
actions and.critical review. The Appendix offers a sketch of key themes and
changes in the successive drafts.
It is an invariable temptation of the analyst's craft to discuss another's
work in terms of how the reviewer would have done the same task. Some of that
tendency will be reflected in my commentary on these drafts--but my purpose is
to use this example to raise some questions about how National Estimates can be
made more useful and accurate. hopefully, these observations will lead toward
ways in which "we can learn to do better in the future", especially on the
specific topic of Cuban foreign policy, which will remain a major problem for
some time to come.
This discussion will have two facets--a substantive critique of what the
estimate drafts said and a commentary on elements of the analytic style and
terminology.
Comments on the Substance
In considering the overall evolution of the 11 drafts, two matters are
most apparent: they were uniformly behind events a A data available to CIA; and,
the essential body of text (called arialysis} remained mc,.tly intact while the
major conclusions changed dramatically. It is also remarkable that the long,
elegantly-written text gave rather little attention to the facts of Cuban
behavior known to the intelligence community and likely to be of enormous
interest and value to policymakers if synthesized and attractively presented
(using maps, charts, graphs as well as prose) ** Detailed comments will cover
four themes: the gap between events and the analysis; the evident "mind-set"
problem; the misunderstanding of the Cuban-Soviet relationship; and the parochialism
which failed to take account of Cuban actions outside Latin America.
'Me Appendix at pp 9-13 summarizes the evolution of each successive draft.
**A marked contrast to the estimate in its emphasis on factual information
about Cuban actions is the_NEAC/OPA paper, Cuba: Promoting Armed Struggle in
South America, September 1981"(PA 81-10372).
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
Estimate Drafts Were Behind Events and Data
Starting in late 1978, CIA increased its coverage and analysis of the
Sandinista guerrillas and their sources of foreign support. In the first five_
months of 1979 (January through May), there were more than 40 special memos,
NID/PDB items, and Latin American Review articles wh'tch described an
increasingly active and effective Cuban support system. On 2 May 1979 a draft
prepared by OPA/NFAC came to this conclusion:
"The Castro regime apparently concluded by at least last fall
that prospects for revolutionary upheaval in Central America
over the next decade or so had markedly improved. . . . As a
result, Cuba has intensified its efforts to unify insurgent
groups not only in Nicaragua--where Cuba has concentrated its
efforts--but in Guatemala and El Salvador as well."
(PA-M-79-10208, 2 May 1979, p '1)
Following the Sandinista victory in July 1979, there were many NID/PDB
items summarizing the extent of Cuban support for the new revolutionary
government; and as early as 22 August 1979, an NFAC memo analyzing Cuban
and Nicaraguan (i.e. FSLN) plans to support Central American guerrillas with
training to begin in Nicaragua. The NID of 24 August 1979 reported Sandinista
plans to train Salvadoran guerrillas and an NFAC/0PA analysis of 27 August 1979
was the first in a series of monthly memos for National Security Adviser
Brzezinski and others which chronicled the expanded Cuban and Nicaraguan support
for the extreme left in the region. From September 1979 to July 1980 there were
about 40 to 50 additional NID items or memos on this topic. In December 1979
there was a very important planning and unification meeting in Havana where
six extreme left groups from El Salvador established a unified command structure
and where Castro w:as, explicit about his 25X1
intentions to help them take power in El Salvador while strengthening the
insurgent groups in Guatemala and leaving Honduras as a "funnel". The public
announcement of the guerrilla command in El Salvador in .ianuary 1980 was
followed by a well-documented increase in Cuban/Nicaraguan support and enormous
increased violence in El Salvador (estimated victims increased from about 800
in 1979 to about 9,000 in 1980.
How did the July 1980 draft describe Cuban policy in the region? Although
stating that "support for violent revolutionary groups in Central America has
increased", the draft emphasized that Cuban policy was "in transition" while
an "encompassing reappraisal" was under way (p 1). Concerning the future, the
draft said, "we expect that Castro will pursue ambivalent and contradictory
policies during the next three to six months, while gradually moving toward a
radical renovation of his regional policy." (p 3)
Concerning the Cuban presence in Nicaragua--which NID and other CIA
products had described as directly supportive of Sandinista consolidation of
power--the-July 1980 draft says: "the Sandinista government also looks to
Cuba as its primary inspiration and benefactor. . . . Between 3,400-4,000
Cubans provide a multitude of services in virtually all areas of public life"
(p 27).
O
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
As the pace of revolutionary violence increased dur'ry the next year,
the estimate drafts came closer to conveying some of the factua, information
1979-80 information but still failed to reflect or present many relevant events
and analyses of 1981. A synthesis of Cuban actions in Latin America (titled
"Cuban Covert Action in Latin America") was prepared by CIA and DIA in 1981
with the Department of State producing a draft in May 1981 very close to that
which was finally released to the public. This report described the synergistic
connection among the different elements of Cuba's subversive apparatus as
follows:
available to CIA, but the gap remained so that an outside reader would at
first presume the author had little more access to information than the
average private scholar.
The June/July 1981 final estimate and analysis had caught up with the
Cuban subversion today is backed by an extensive secret intelli-
gence and training apparatus, modern military forces and a large
and sophisticated propaganda network. Utilizing agents and
contacts nurtured for more than twenty years, the Castro government
is providing ideological and military training and material and
propaganda support to numerous violent groups, often several in
one country. ("Cuba's Renewed Support for Revolutionary Violence",
(U), 14 Decemter 1981, p 1)
Except for El Salvador and Guatemala, neither .he factual information nor
the sense for the broad pattern of Cuban actions, as described in this intelli-
gence community product in the spring of 1981, was in tie final estimate. Nor
for that matttr was there more than an occasional paren'1ietical reference to a
series of violent attacks st?ned by Cuban-svppcirted groins which followed the
tough rhetoric of the new US adr.,inistra'.ion. These eve-its in(1 uded, among
others:
Costa Rica - armed attack of US Embassy guards (Mar. 81).
- four terrorists imachine-gun police veh?--cle (Apr. 81).
- terrorists murder three policemen and a taxi driver;
links to Argentine Tupamaros, Uruguan Montoneros,
Colombian M-19 and Cuba directly are firund (June 81).
Honduras -
- infiltration of terrorists from Nicaragua with target
being the Guatemalan Embassy staff (July 81).
bombing of Honduras parliament; seizing of Honduran
airliner; increase in terrorist infiltration from Cuban
training camps.
Colombia/Panama - murder.of Mr. Chester Bitterman by the M-19
guerrilla group; sending of 90 Colombian M-19
guerrillas from Cuban training camps, through
Panama, to attack Colombia.
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRETI 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
possible assassination of President Roldos claimed by
Colombian M-19 guerrillas in retaliation for his having
turned captured terrorists over to the Colombian government.
The estimate would have been more complete if it had examined the
implications of these and other events.
The Mind-Set Problem
The radar operator at Pearl Harbor then, or NORAD today, who doesn't use
data that disagrees with his perspective is the classic example of the mind-set
problem long identified as a cardinal danger in the world of intelligence. Of
course, as one colleague has noted to me, it is often hard to distinguish
between a "mind-set" (or a "construct" when used positively) and wisdom.
It is absolutely clear that the primary author of these drafts came to the
task with a deep, rich,and rather elaborate sense of whc Castro thinks he is
and what motivates his behavior. The drafter also had eery definite beliefs
about how Castro viewed the US and what Cuban domestic failures implied for its
foreign actions. However, little of the beautifully-writer text on these
themes informs the reader whet evidence the author 4s t.'ing or even how he
reached his concl,.siors.
On a number of issuss that throughout the 11 di aft rerr,ii, the ostensible
keys to the interpretation and prediction of Cuban acti ir,s i,' Latin America,
the judgments of the draft change completely as the "mind=set" seems to melt
away--slowly over one year. Each elem,?nt of this "mind-set" will be discussed
briefly.
The July 1980 draft sees "different Cuban actions if Castro does not reach
some accommodation with the US" (p 5). Not until draft number seven (11 May 81)
does this interest in "accommodation" facie away as a much sharper and more
confrontational view of Castro is presented. My question: in the light of the
failed normalization efforts from 1974 to 1979: wfy did the drafter think that
"accommodation". was on Castro's mind?
A related point is that in the early drafts, Castro's concerns about possible
US reactions restrict and limit his actions. But by the final estimate, this
changes completely, as we are told that,"In his logic, militancy and defiance
will be seen as the best course for keeping the US at bay and for creating
sufficient revolutionary siblings in the region to insulate him over time from
US pressures" (July 81, p 14). But, why the surprise--offensive action has been
Castro's method for two decades?
Initially, Cuban domestic failures and frustrations were seen as oartially
distracting Castro and limiting his international subversion. But the final
estimate reverses this judgment: "Castro's emphasis on support of insurgencies
also reflects his frustration over domestic problems" (June 81, p 3). This is,
of course,-a perennial issue in foreign policy analysis--do domestic problems
limit leaders' actions abroad or encourage them. to use the international arena
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRETI 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
as a diversion? Either can be true at different times. But because the issue
is so basic and comes up so often, the CIA might have been expected to have
presented a logical case for one interpretation or another--these drafts do not.
Another perennial theme--especially in dealing with dictatorships--is the
question of "pragmatic" or "hardline" factions' impact on foreign policy. The
July 1980 draft discusses the "rise of the hardliners" which is said to have
begun with the appointment of a tougher :Interior Minister in December 1979
when "the first intimations of a return to Stalinism were heard" (p 11). The
draft also notes "though of late, Castro has been more attracted to the hardline
school of thought, he has taken no steps yet decisively to upset the balance
of power among his various advisors" (p '13).
Aside from admiration for what must be incredibly good sources for such
difficult judgments, the reader is led to wonder whether it was the "pragmatists"
who counselled Castro to send tens of thousands of Cuban troops to Africa and the
Middle East during the 1970s? The final estimate continues to view the "rise"
of the hardliners as an important causal factor (July 81, p 4-5) but does note,
"A cardinal tenet of the Castro government froa- its inception has been that
support for revolutionaries elsewhere in Latin America is a right and duty of the
Cuban revolution" (p 5). That is correct and this fact along with the
aggressive behavior of Castro over many years or three continents calls into
considerable question the operational significance of tte rather elaborate
"hardliner"/"pragmatist" discussion.
Contrary to all the evidence, the early drafts ind"cated that Castro's
desire to protect the revolutionary victory in Nicaragur would make him more
cautious. By?the last draft, this presumption is corrected by noting that
revue t onary success in El Salvador and Guatemala will help defend the Nicaraguan
gains. This insight was about two years behind CIA datf3 and judgments.
It is worth noting that the "mind-set" about Cuba did not permit any
serious discussion of Castro's past successes in neutralizing or ignoring US
opposition to his actions. Nor was there any consideration of Cuban aggression
in Africa/Middle East during the same time that the three US administrations
sought to "normalize relations".
Parochial Failure to Take Account of Cuban Actions in Africa and the Middle East
All the drafts display an astonishing parochialism in failing to present
and take account of the Cuban subversive and military actions in Africa and the
Middle East from 1970 to the present. These are of interest in understanding
whether Cuba is moderate or revolutionary. These actions are directly relevant
to an analysis of expanded Cuban actions in Latin America in at least the
following ways:
- better than Castro's rhetoric or academic literature in Latin A.morican
studies, Cuban activities during the 1970s demonstrated a strong commit-
mentto revolutionary warfare and operational cuordinatien with the
Soviet Union;
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
- the evolution of Cuban activities In Africa--from the beginning of
subversion in 1962 to tens of thousands of troops in action--might
offer some lessons for Latin America;
- Cuban intelligence, propaganda and military skills and assets were
greatly increased by the Middle Eastern/African experiences of the
1970s ;
- Cuba's use of various third-country terrorist groups and its capacity
to use proxies and cut-outs were expanded enormously in the Middle
Eastern/African arenas with the effects visible in its careful, concealed
and mostly competent handling of its new Nicaraguan partner;
- Cuba, in working with the more experienced secret police of East Germany
and the Soviet Union in building internal security systems, establishing
reliable military cadre and military internal security networks in Africa,
probably increased its skills in c:onsolidatinq revolutions. These are now
being applied in Latin America.
Cuban-Soviet Relationship Presented in an Over-Complicated Way
The final estimate states simply "The Soviet-supported Cuban challenge to
US regional interests will continue to be formidable" (J~_ine 81, p 2). But this
is in contrast to most of the drafts where an overly complex and convoluted
treatment of the Soviet-Cuban relationship essentially creates the impression
in the reader of a difference in purposes or of Soviet restraint on Castro
(for example, the 5 November 1980 draft argues that Moscow will restrain Castro
in South America).
The first draft makes this rather remarkable and, to me, unintelligible
statement:
"As Cuba's dependence on the USSR and willingness to perform proxy
services increased during the 1970s, the relationship with Moscow
actually declined as a burden on his maneuverability." (July 1980, p 8)
This was somewhat contradicted by the news on the following page of the same
draft that Cuban reaction to the Soviet invasion of Afghanistan in December 1979
"confirmed the true nature of Cuban subservience to the USSR". (July 1980, p 9)
Had the author not heard of the events of 1968 after the invasion of
Czechoslovakia--the Soviet cutoff of oil to Cuba and the takeover of the DGI by
the KGB or any other number of events and actions which clearly showed that, Castro
could be controlled and ordered when necessary? None of this relevant background
is mentioned in the discussion of Cuban-USSR relations despite being amply
documented both in the open literature and earlier CIA studies of Cuba. The
final estimate says, "The Soviet Union and Cuba share similar objectives regarding
Latin America" (July 1981, p 9), and the strong evidence of Soviet support for
Cuban tactics in Central America was well in hand by tb-e summer of 1980 when
the writing -was still so ambiguous--despite the preceding decade of collaboration.
Why? -
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET
Approved For Release 2006/01 - 00269RO01100100007-1
Comr^ents on Analytic Style and Terminolo Y
Much of the CIA writing which I have seen to date seems to focus on
judgments about the mood, expectations, or potential reactions of foreign
leaders or governments while offering little infcrmation about the actions
of those same leaders or governments.
This series of drafts reflected this 'tendency in twc ways--a focus on
the moods and beliefs of Castro and an almost theatrical hypothetical dialogue
between "hardliners" and "pragmatists".
The Castro centeredness is especially perplexing because when positive
things happen, he is the initiator, but when negative or aggressive actions
are taken, he is being pushed or pulled by the "hardliners". For example,
the final draft says, "Castro himself has been attracted to more active
support.fbr regional revolutionaries, in good measure because of his frustration
and anger over domestic and foreign policy setbacks". (July 1981, p 4)
We also learn that "Castro no longer has any illusions that Torrijos
will become a dependable ally" (July 1981, p 8) which prompts the question--
did he ever think that or did the author believe Castro once expected that?
In a word, the unremitting focus on Castro's hopes, fears, and beliefs, all
of which are presumed to change dramatically over the course of the 11 drafts,
leads to the question--do these statements about Castro derive from data (so
it would seem to a new reader or outsider) or do they serve as a literary
device for conveying the author's changing opinions?
This question applies even more to the remarkable use of the "pragmatist"
vs "hardliner" theme where the author(s express bel ieft. about what the
hardliners are telling Castro:
- "We believe the hardliners hold that armed >truggle is the only
effective way to achieve revolutionary change.
- "They probably are convinced that the correiaticn of forces now
favors the communist camp."
-"They probably believe US threats will strengthen domestic .support
for the Castro regime." (July 1981, p 4)
Last, is an observation on what seems to me a pattern of terminology that
conceals the human and political meaning of what communist revolutionary warfare
brings. Through all 11 drafts we read about:
- "sibling revolutionary regimes" not pro-Cuban dictatorships,
- "radical renovation of his regional policy'', not increased terrorism
and guerrilla warfare.-
"the romantic and nationalist" phase of Curran foreign policy--not
subversion and.aggression against the struggling democracies--
Venezuela, Colombia, or armed landings in Guatemala, Panama, etc.
SECRET/+
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET/I 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
- "enticing new opportunities for expanding its regional role"
not a chance to bring down the Central Airerican governments. sti
a Granted that terminology is Somieteba~keda totalitarian state atnackstitsxpress -
clearly what happens when a
neighbors.
SECRET4 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET 25X1
Approved For Release 2006/ B00269R001100100007-1
APPENDIX -- Hi hli hts and Evolution of the Cuba/Latin America WE
-- 8 Success ve Drafts
Ju y 980 to Ju y 1 =
1/ 18 July 1980 -- 1st draft
- Highlights Cuban "reappraisal" of policy for international and domestic
reasons.
- Predicts "ambivalent", contradictory policies over next 3-6 months, but
toward more radical.
- Cites "growing fear of US" as a somewhat constraining factor.
- If Cuba does not reach "accommodation" with the JS by early or mid `81,
then would certainly go to hard line.
- Generally, characterizes Cuba's foreign policy as being "in transition",
things still tentative.
- Fear of US said to be a "much larger factor constraining Cuban policy".
- Makes much of the "hardlines" vs "pragmatists" in Cuban circles.
- Stresses, as a motivation, the Castro regime's setbacks at home and abroad.
- Much historical background.
- Has some discussion of South America that does not "fit" rest of discussion.
2/ 23 September 1980
- Still says Cuban policy is "being reappraised" but notes radical changes
that have emerged already.
- Support to leftists has "intensified" and that commitment will continue
to grow "at least at a guarded pace"..
- Still attributes much importance to Castro's desire to avoid a clash between
his"hardliners" and "pragmatists".
- Strongly suggests that a more radical course would isolate Havana in
Latin America and that this a major deterrent for Castro.
- Avoids (as does first draft) any real judgment on now likely the "radical"
option is.
- Reordering of paragraphs, but no apparent leap forward in judgments.
D
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 SECRET
Approved For Release 2006/
1/30 : CIA-RDP8
3/ 23 October 1980
- Degree of commitment to revolutionaries still said to be likely to grow
"at least at a guarded pace"; some effort to distinguish more neatly
between "prudent" support of left and so-called "reckless" approach
(i.e., radical policy).
- "Reckless" option -- now defined as "aggressive and impulsive", and still
hinging on failure to reach accommodation with US.
- Rationale for this "reckless" approach is spelled out here for first time
(i.e., to tie up US so that it cannot hit directly at Castro).
- Both drafts mention that Cuba wants armed revolt in Honduras.
- New draft says Havana's militant line helped cause split among Honduran
leftists.
- This draft more explicitly cites course of US-Cuban relations as a factor
in determining whether Castro adopts a "reckless" co,irse or not.
- Gives odds (1 in 5) for likelihood of radical, -eckless" course.
- In Key Judgments, gives somewhat less prominence to factors that could
constrain Castro.
4/ 5 November 1980
- In Key Judgments -- adds a constraint. Cuba will hesitate to go for
broke in El Salvador and Guatemala, because to oo sc would endanger
principal aim -- which is to safeguard revolution in Nicaragua.
- Also says Moscow will restrain Castro in South Amerca, where Soviets
have important equities they don't want threatened.
- Highlights status of US-Cuban relations as a major factor in determining
Cuba's regional actions.
Stresses that Havana has unique advantages over US in exploiting problems
in region (nationalism, Latin suspicions of US, etc.).
- Changing wording on ascendancy of "hardliners" reflects uncertainty over
just how important a factor that phenomenon is.
- Highlights the fact that El Salvador and Guatemala are the "priority
targets".
- Terms-describing Cuban presence in Nicaragua and the importance of
Nicaragua revolution to Cuba are made more dramatic.
- Updated to take account of Seaga win in Jamaica -- this is interpreted as
contributing to Castro's-reliance on violent tactics.
SECRET/
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
- Section on "Need for Mexican Support" is condensed.
- As in previous two drafts, difficulty is in assessing whether paper
really sees "reckless" or "prudent" policy as more likely (despite Key
Judgments).
- Second and third drafts have revised sections on "Conclusions and
Implications for the US" -- not spelled out this way in first draft; this
device attempts to spell out more clearly than 1st draft, the effect of
US ties on Cuban policymaking.
- New sentence: "Castro would readily sacrifice any improvement of US
ties to take advantage of any opportunity that might arise." (This, then,
calls into question the validity of saying that LIS ties are vitally
important to Cuban strategy in the first place.)
- Throughout three drafts, hard to tell if drafter rea ly believes US
relations are causal or not -- community differs on this and drafter may
be seeking ways to accommodate.
- Hindrances to US room for maneuver are now couched more in terms of Latin
"hangups" rather than in terms of residual Cuban leverage against US.
5/ 17 December 1980
- Language perpetuates the problem of determining whether the draft in
fact is predicting an all-out "reckless" course by Castro in Central
America or not ("prudent guarded pace", etc.).
- Highli hts reluctance of region's "progressive" governments (Mexico,
Panama) to break with JRG, e.g., as a restraint on Castro.
- Paragraph 73 -- The degree to which Castro feels (a'; some point) trapped
and threatened may be one key (in pushing him to a reckless course).
Former drafts said this could be the key.
- Drops the caption, "Implication for the US".
Another "key" added is Castro's perce tp ion of the likelihood of a
forceful US response.
6/ 10 March 1981
- New format -- "Key Judgments and Outlook" precede text, and their tone is
different.
- First part of Key Judgments now stresses importance (as a motivating force
for Castro) of his perception of increased opportunities in region, con-
trasted with earlier (seeming) primacy of his setbacks at home and abroad
(those -now come second).
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRET 25X1
Approved For Release 2006/0130 - CIA-R__ 00269R001100100007-1
- Also highlights supposed Cuban fears of US countermeasures more than
earlier drafts.
- Similarly, now highlights Castro relationship with Moscow and effects
of his actions on relations with USSR -- stresses Moscow as a constraint
on Havana.
- Outlook, country-by-country, is attached to Key Judgments.
- Returns to an earlier draft's recitation of Cuban advantages and US
disadvantages in region and adds to the list.
- Body of text largely similar to earlier drafts.
7/ 11 May 1981
- Format changes again -- now leads with "Summary Conclusions", then Key
Judgments.
- Now the Sandinista success of '79 along with Castro's setbacks said to
motivate him.
- US policy said to make Castro's choices tougher.
- Says Castro can perhaps be slowed, but not deterrec.
- Now, USSR is said to "share" Castro's goal of undermining US influence;
but Moscow, if influenced (undescribed US pressure; might act to brake
Castro.
- Also says US policies in region will not affect area's inherent
instability -- but could "raise costs" for Cuba, USSR.
- Key Judgments now lead off with statement on "Castro's persistent objectives"
in Latin America -- a generally sharper, confrontational characterization of
him.
- Also features USSR angle more prominently than before -- but is wishy-washy
influence
inahemisphere~0nal policy --
on extent to which Soviets
though suggests he has considerable - Generally, much lengthier coverage of Cuba-USSR, and USSR interests in
Latin America.
- Otherwise, text itself largely similar to previous drafts.
8/ 15 May 1981
- About the same as 11 May draft;
9/ 21 May 1981
- About the same.
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRET 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
10/ 15 June 1981
- Now estimate is divided in to separate pieces -- "The Estimate" and
"Supporting Analysis" (to be issed later in July 1981).
"The Estimate"
- Introductory Statement -- appears to be newly written from the ground
up -- though in some ways synthesizing from past drafts.
Highlights Cuba's revised foreign policy, improbability that Castro
will make any concessions to US.
Highlights Soviet support for region's leftists and Castro's
"militancy"
- Is followed by modified version of most recent key Judgments again
highlighting Castro s reaction to greater opportunities in region, as
well as his reaction to setbacks.
- New paragraph -- specifically rules out use of luban combat forces in
Latin America "in the near term".
- Now adds to list of Cuba's strengths, its "massive" economic and military
assistance from USSR.
11/ 9 July 1981
- By and large repeats body of previous version of text. (International
setting. . . . Cuban setting. . . . etc.)
SECRET)
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
15 January 1982
Review of Nicaraguan Analysis Since January 1979
Your request for comments on the Nicaraguan analysis cites the realistic
sub-
CIA memo 2 may
asks: picture as earlier Nicaragua.".
sequent t estimates
The basis for these comments includes a review and summary of the main
items of analysis produced by CIA both in the regular reporting series such as
the NID and PDB (see Appendix A, pp 1-4) and special memoranda sent in more
informal channels (see Appendix B, pp 5-8). This production record demonstrates
that CIA performed well in reporting many aspects of the developing situation.
What is remarkable in the light of the information represented by this reporting
effort is that the Carter administration policymakers tended to ignore events and
the CIA reports until major crises forces some type of response. This is most
clearly illustrated by the following:
- In mid-June 1979, Presidential Review Memorandum 46 was prepared
by State and included as an appendix the CTA memo of 2 May 1979.
However, virtually nothing in the text for the President conveyed
any real sense for the degree of Cuban and other international
communist involvement with the Sandinistas.
- Starting as soon as one month after the 19 July 1979 Sandinista
victory--which had confirmed the previous reporting--CIA reported
that the Sandinistas and Cubans were aiding guerrillas from
El Salvador and would step up the pace of attack; the Carter policy-
makers failed to even consider new policies or approaches for the
obvious next target country until the December 1979 meeting in
Havana unified six extreme left groups and then led to massive
demonstrations in E1 Salvador.
Many other examples could be cited--including the unwillingness of the former
US Ambassador in Nicaragua to face up to the real facts of Sandinista actions, to
nopolize power and the former US Ambassador in El Salvador who simply turned
dray all the CIA information about the guerrilla buildup which was followed :by
the 10 January 1981 offensive with the comforting illusion that they were "only
juvenile delinquents" with no real combat capacities. this still leaves two questions: how good was the CIA analysis and could the
ir:telligence community have done more to persuade the policymakers to pay attention?
On the question of CIA performance (based on a reading of the material), there were
three very good elements, but each was associated with less effective aspects; and
there were three large areas of omission which are only now being remedied.
Followi,pg the failure of the summer/fall 1978 Sandinista insurrection and
the intense US mediation efforts, CIA performed well in recognizing that a large
crisis was impending. As early as January 1979 the MID warned about this.
SECRET
25X1
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
However, when a revolutionary situation Is escalating, the primary intelligence
question is (or usually ought to be) when will it seriously endanger the
threatened government? As late as May 1979, the Agency slid Somoza would probably
last till the end of his term in 198;. Probably a bit of "group-think' was at
work and might have been remedied by a conscious and specific effort to explore
more pessimistic possibilities-1-along with Indicators. As with th? Shah,
Somoza was'not defeated; his army collapsed after morale .as shattered
by a combination of international and internal pressures. It seems that mare
analysts have difficulty empathizing with the life and death character of
revolutionary warfare and the effects on threatened elites of their personal fears
and expectations. In contrast, communist political leadership (with Sun Tzu) has
always emphasized breaking the will of their enemies by false offers of leniency
for collaboration or early surrender.
Reporting on Cuban support for and influence with the Sandinistas was very
good and consistently improved as additional efforts were made. However, after
the victory, despite a great deal of information, there was not enough realistic
assessment of Marxist-Leninist techniques and actions to consolidate power. This
was true despite the virtual repetition of the Cuban process without the mass
executions and trials due to Castro's sound advice that deception be used to
neutralize the US and have the West finance the new regime.
Also first rate was the reporting on Cuban and Sandinista plans and actions
in Nicaragua immediately after the victory for the purpose of supporting the
El Salvador guerrillas. However, this was not translated into any hard-headed
assessment of how quickly the guerrilla threat there might grow later in 1979 and
1980. Again, perhaps a case of "group think" since most analysts at first seemed
to think it would take two to three years for the buildup to reach the level
actually attained in about one year.
From information seer;, there seemed to be four key om;ssions in reporting
and analysis:
- Pouch too little attention was given to the dramatically new level
of non-communist support for the revolutionary left in the region--
including Mexico--one overall analysis in June 1980 of a "new
activism", but no focus on Central America de pi to the key role played
by Mexico in undermining Somoza in May 1979; the same holds for the
lack of informed attention to the financ aI, practical and propaganda
support being given by the many Social Democratic groups (one report, Jan 81
- In spite of long-established patterns of collaboration among the
Soviet Union, Cuba and radical Arab forces such as Libya, the PLO,
PFLP, DFLP. . . and many items of public and intelligence information,
too little attention, thought and analysis was given to their role
in support of the Central American extreme left--also a dramatic change
.in this region.
SECRET)
Approved For Release 2006/01/30 : CIA-RDP86B00269R001100100007-1
 SECRET 25X1
Approved For Release 2006/0113U : - 00269R001100100007-1
- Also, too little information has been collected and reported on
the international sources of support for anti-communist moderates e.g.
the free trade unions, Christian Democratic Parties, and countries
like Venezuela--few analysts and policymakers seem to know much
about this important group of participants.
= International links among the extreme right groups and their
connections to exile funding sources in the US have apparently
received hardly any effort.
Efforts are now being made to fill in information on the first two gaps.
DDI/ALA has a good report on the Palestinian extremists in draft; an IIM on
Mexican actions in Central America is being written and an IIM on the trans-
national forces for and against the extreme left is scheduled. However, this
is about three years after these groups and countries became publicly (as well
as secretly engaged in a highly significant way. Underlying intelligence
issues about Mexico and the social democratic groups have not yet been addressed
(to my knowledge). These include: the extent of Cuban/Soviet penetration of
the Mexican foreign ministry, the PRI_and COPPAL, and similar questions about
the social democratic groups and the German political action foundations.
Three Practical Suggestions
No one is or can be a prophet in looking ahead at fast-moving revolutionary
warfare situations such as those in the Central America/Mexico region. 'But,'
analysts could be encouraged to stretch their minds a bit more into the future
on a probabilistic basis and do three things:
1/ In cases like El Salvador, Guatemala, Honduras, Costa Rica --
where there is clear evidence of a major attack on the government is
in progress or is coming -- analysts should provide for the
expression of a consensus analytic view and one or two more
.pessimistic alternative forecasts with indicators by which to 'Judge
their probability. The alternatives should be seen as an exercise--
"
this is now it could go if we are more pessimistic"--not a
prediction for which the analyst will be held accountable..
2/ Role play the enemy forces. For example, in E1 Salvador, "i f I were
the FMLN commander, here is the way I would try to bring down'the
government--internally and internationally--in the next 6-9-.months".
Let several analysts do this independently as a shcrt (one-man day)
exercise, compare notes, and then sketch out a hypothetical plan
along with indicators of whether anything like that Is being tried.
Very important here is to project ahead on the basis of the inter-
national political calendar as well as the internal situation.
Again, this should be seen as an exercise in creative intuition not
predictions which come back to haunt the analyst.
SECRET
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
3/ Occasionally, create a lively person-to-person debate in a seminar
setting (2-3 hours) where a small group with some relevant background
but no current analytic responsibilities in the area receive in
advance, firs the facts In the form of chronologies, data, TDs, etc.,
second, a few major analytic products deriving from that data base-
and then use the seminar to ask questions and make suggestions.
Members of PFIAB or the Senior Review Panel, plus some DDO personnel
and the NIOs might be good participants in such a bi-monthly session
on a few key issues.
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
 25X1 Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1
Next 7 Page(s) In Document Exempt
Approved For Release 2006/01/30 : CIA-RDP86B00269RO01100100007-1