INTELLIGENCE AND THE PREDICTABILITY OF MAJOR INTERNATIONAL EVENTS

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP91M00696R000800100003-5
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 5
• Document Creation Date: December 15, 2016
• Document Release Date: August 20, 2004
• Sequence Number: 3
• Content Type: REPORT

Approved Forelease 2004/09/03 : CIA-RDP91 M006900800100003-5
"Sensitive Issues-and Allegations"
No. 10
Intelligence and the Predictability
of Major International Events
National intelligence is far more than the prediction
of specific events. As intelligence has evolved in the
years since World War II, it has become an integral con-
tribution to the process by which national policy is shaped
in,today's complicated, interdependent, and dangerous world.
In this context, I believe that our responsibility in
the area of "prediction" is best defined as the prevention
of surprise. In fact, for intelligence to make its primary
goal simply the prediction of specific events would in fact
be a disservice to the policy officer. We would often be
wrong, and almost never exactly right. For we would be
claiming for intelligence a degree of precision in defining
human intentions and in choosing among contingencies that
is not within the power of any human'individual or group
that we know.
But we should be able to prevent the policy officer's
being surprised by any event of major importance. If he is
surprised, we have failed, either because our judgments were
faulty or because we did not adequately communicate the
degree of, our concern. We succeeded on both counts regard-
ing the Middle East War in 1967; we missed on both counts
in the Middle East War in 1973.
What can the policy officer reasonably expect of us?
He can expect that we will put him in the context of events
as they occur; that we will help him understand the dynamics
of a situation; that we will lay out a range of possible out-
comes, especially those that damage US interests or present
an opportunity to the US; that we will seek by further col-
lection and analysis to narrow this range, to reduce many
possibilities to a few, and to rank them; and that we will
warn him at the earliest possible time of any increase in
the odds for a serious crisis. But we must not cry wolf too
often. Sooner or later one is sure to be-right that way,
but the policy officer will have long since ceased to listen.
Approved. For Release 2004/09/03 : CIA-RDP91 M00696R000800100003-5
 Approved Forlease 2004/09/03: CIA-RDP91 M0069ft00800100003-5
The India-Pakistan War of December 1971 is a good
example of what intelligence at its best can do. At no
point did we forecast a specific event on a specific
date, but we were able in April 1971 to forecast the
nature of Indian military support for the Bengalees, and
by June to warn that events were moving toward war.
Policy officers were fully informed of the role of the
Soviets and Chinese, of the high. probability of Indian
victory and East Pakistani independence, and of the inter-
national consequences that might flow from the weakness
of the new state of Bangladesh.
Of course, there are occasions, unfortunately rare,
when we can say correctly that event A will occur on
date B, as with the Turkish invasion of Cyprus on July 20
last year. In general, however, the more we can draw a
prediction from physical events or evidence, the more
confident we can be. When, in early 1972, crack North
Vietnamese reserve divisions physically moved to the Viet-
nam Demilitarized Zone, we correctly saw this as prepara-
tion for a major offensive. (Yet it still took a political
decision, several weeks later than we expected it, before
that offensive was launched.)
Because construction takes time we can project accu-
rately from physical evidence Soviet ICBM strength as much
as two years hence. Longer range projections are neces-
sarily less accurate. In fact, hindsight shows that we
were consistently high in the late 1950s and early 1960s--
lacking hard evidence, we had to extrapolate short range
trends. We were too low in the late 1960s because we could
see little military need for Moscow to go for a larger force,
and expected to see more emphasis on a technologically better
force at that time. We did not give enough weight to Moscow's
political need to match--and in some respects to exceed--US
strategic forces in size as soon as possible. When we are
projecting ten years or more out beyond what clear evidence
will allow--as we must do to support US military planning--
we are entering periods for which the Soviets themselves have
not made firm plans and decisions.
Even when physical evidence is available, its meaning
is usually ambivalent. Our experience is that major inter-
national events usually stem from political. decisions, and
these in turn from personal. relationships within groups of
Approved. For Release 2004/09/03 : CIA-RDP91 M00696R000800100003-5
 Approved For elease 2004/09/03 : CIA-RDP91 M0069000800100003-5
leaders. This human factor can affect our forecasts in
many ways. Here are some examples:
The Last-minute Decision. National leaders.
often take contingency preparations for action,
reserving as long as possible the final deci-
sion whether to act. When the Soviet leaders
took alarm at the direction Czechoslovakia was
heading in the spring and summer of 1968, they
sent forces toward and into Czechoslovakia to
conduct "exercises." We reported throughout
the summer that Moscow had assembled the forces
necessary for military intervention. In late
July and early August, however, a series of
indecisive meetings among the Soviet leaders,
and between them and the Czechs, seemed to take
some of the urgency out of the threat, and it
now appears that the final decision to "go" was
taken only a few hours before the invasion was
launched on 20 August. (We understand from
later information that the Politburo vote was
close.) We had warned the US Government that
preparations had been made, and that at least
some of the Soviet leadership were in favor of
intervention. But we could not say whether
they would in fact intervene--they did not know
themselves.
The Coup Plot. Successful coups are usually made
by small, tight-cliques of military troop command-
ers, operating in total secrecy. Intelligence can
virtually always identify a situation that is rip-
ening for a coup--discredited leadership, faction-
alism, disorder, economic disruption. It can
usually identify interest groups that would profit
from a coup. It can often identify likely plotters
and sometimes penetrate their circle. It can some-
times distinguish the marginal plotters from the
real "heavies." But it can only rarely predict with
confidence precisely who will act, how and when.
Nonetheless, more often than not, it has been able
to give adequate warning that someone is likely to
take action.
Approved For Release 2004/09/03 : CIA-RDP91 M00696R000800100003-5
 Approved ForVelease 2004/09/03 : CIA-RDP91 M0069000800100003-5
The Fortuitous Opportunity. Plotters may plot
for years and never act, perhaps because they
are inveterate plotters and not doers, perhaps
because everyone--and especially the local
security service--knows they are plotting, or
perhaps because they can never find the right
opportunity. On the other hand, the desire
for action may fester for years in the minds of
a plotting group, unknown to anyone else, and
when an opportunity comes they will seize it.
In 1958 an Iraqi brigade was being transferred
from one frontier to another, passing through
Baghdad. Because it was on active service it
carried live ammunition, which the government
wisely did not permit to units permanently sta-
tioned in the capital. A group of officers,
who had long sought such an opportunity, took
over command, diverted the brigade, and over-
threw the government. There was, obviously, no
intelligence warning, although analysts had long
perceived that the basic situation was unhealthy.
The Self-defeating Prediction. One can find on
the intelligence record many warnings of events
that never took place because the warning stimu-
lated policy action to forestall them. This, of
course, is really intelligence doing its job, and
we are not about to wring our hands because the
situation of which we warned has not come to pass
for this reason.
This Committee has called sharp attention to the intelli-
gence community's performance in the 1973 Arab-Israeli war.
You have available a detailed study that we ourselves under-
took about what went wrong with our judgments regarding the
imminence of hostilities at that time. No one is going to
argue that there was not a serious failure.
What is important now, however, is what we have learned
and what we have done as a result of this chastening experi-
ence. Let me give a few examples in this category:
-- We have improved informal substantive give-and-
take among expert analysts in the intelligence
community. The National Intelligence Officer
Approved. For Release 2004/09/0: CIA-RDP91 M00696R000800100003-5
 Approved Forelease 2004/09/03 : CIA-RDP91 M0069000800100003-5
for the Middle East calls analysts together
whenever the situation appears to be heating
up,-and explores with them facts, perceptions,
and judgments.
We have turned to more systematic techniques
of expressing analysts' judgments on the prob-
ability of a given event--such as the outbreak
of hostilities--occurring within a given time
period. The technique we are using also dis-
plays for the policy officer the range of
analysts' individual judgments on these issues.
We have revamped our approach to estimating
Middle East military developments at the na-
tional intelligence level. We are now not
only identifying the strengths of the major
antagonists but also trying to envisage alter-
native ways in which a war might develop and
be played out. This has been especially help-
ful to contingency planners.
Nor have we been shy about re-entering the
estimative field in the Middle East. Major
estimates of "next steps" in the situation
were produced at critical points during the
past year. Our customers have indicated that
these were generally well received and regarded
as. useful.
In conclusion, I can assure the Committee that we in
the intelligence community are not only aware of our own
human limitations and of the constraints imposed upon us
by the very nature of the human affairs with which we deal,
but are seeking as actively and as imaginatively as we can
to assure that US policy officers are not taken by surprise
and that they have the fullest possible understanding of the
context in which major international events take place.
Approved .For Release 2004/09/03 : CIA-RDP91 M00696R000800100003-5