As the crisis proceeded, CIA documents continued to paint the
bleakest picture of Gorbachev's prospects. "The Soviet Cauldron," a
fifteen-page memorandum prepared for the nsc in April 1991 by the
director of the Office of Soviet Analysis, George Kolt, stated that
"Gorbachev's credibility has sunk to near zero." A May 1991
Intelligence Assessment published by the Directorate of Intelligence,
"Gorbachev's Future" stated that within a year "a major shift of
power to the republics will have occurred unless it has been blocked
by a traditionalist coup."

In retrospect, the CIA was most prescient in anticipating events.
Intelligence community assessments during 1989-91, and the CIA's in
particular, were remarkable not only because the agency was willing
to assert that the communist order was finished--a view that would
have been unthinkable just a few years earlier--but also because of
its sophistication in interpreting the interplay of factors that were
at work, and how these factors constrained the range of potential
outcomes. Specifically, the CIA:

*Repeatedly mentioned a coup as a serious possibility, because the
economy was in shambles and order was breaking down.

*Noted that, if a coup were launched, it would be cloaked in the form
of a "committee of national salvation" that would claim it was
imposing emergency measures to restore order and preserve the moves
toward democratization and economic reform in the long term (almost
exactly the formulation used by the hardliners in announcing their
attempted takeover in August 1991).

*Predicted, however, that the long-term prospects of a coup were not
good, that events in the USSR were being driven by tidal forces. The
wave of nationalism and decay of public support for the Bolshevik
regime could not be reversed; a coup would not alleviate the economic
failure that was in large part responsible for events. Even if
hardliners did manage to seize power temporarily, they would not be
able to consolidate control.

In addition, the intelligence analysis laid down markers--that is, it
presented facts that a reader would have to refute in order to reject
the analysis presented in the assessment. For example, "The Soviet
Cauldron" noted four months before the August coup attempt that:

*Hardliners had raised the possibility of using the military to
restore order in public statements;

*Military officers inclined to support democratization had been moved
out of important posts or retired; and

*The military and security organizations had demonstrated the
logistics and capability of moving large numbers of troops into
Moscow on short notice and establishing a command structure to
control them.

In order to believe that the hardliners would not launch a coup,
therefore, one would need to explain why they would have taken such
potentially risky or costly steps.

Although the record shows that the intelligence community was aware
of the untenable situation in the Soviet Union and gave ample warning
that Gorbachev was in jeopardy, former high-level U.S. officials
disagree as to how aware they were of these warnings. Some officials
at the middle levels of the National Security Council, such as Rice,
knew of both the range of opinion within the intelligence community
and the degree to which they agreed Gorbachev was in trouble. Other
officials we interviewed who had less routine contact with
intelligence analysts tracking the political situation in the Soviet
Union were not aware of these differences, nor that the CIA had an
especially pessimistic assessment of Gorbachev's prospects.

At the highest levels, Brent Scowcroft, President Bush's national
security advisor, does not recall receiving warning that was
sufficiently precise to support action on the part of the United
States. He believes that, if such warnings were given, they were
"lost in the fog" of the volume on information that passed by on a
daily basis. However, Gates recalls differently, and cites three
separate events that would indicate these warnings reached their
target.

The establishment of the "contingency planning group" was one. The
second, according to Gates, was a memo he sent personally to
President Bush in the summer of 1990 citing the CIA's warnings of the
deteriorating Soviet political situation; Gates described Gorbachev
as the "Soviet Moses"--that is, the leader who would take the Soviet
people to the promised land, but would not himself survive the
Exodus. The final indicator was another memo that summer in which,
Gates maintains, he told Bush that the United States should "stop
knocking Yeltsin" because we "would eventually be facing him on the
other side of the table." All of these actions, Gates asserts, were
taken in light of the CIA's briefings and reports.

In any case, it is clear that the Bush administration chose to stand
by Gorbachev in spite of the intelligence that argued his future was
limited, not because U.S. intelligence suggested that this was a safe
course to take. If a failure occurred, it was not an intelligence
failure but a policy failure, and it may not have even been that. The
administration had three powerful reasons for deciding to continue
its support of Gorbachev, despite the intelligence at hand that
suggested his position was precarious.

First, the Gorbachev regime had supported or acquiesced in many U.S.
objectives--the reunification of Germany, the dissolution of the
Warsaw Pact, withdrawal from Afghanistan, nuclear and conventional
arms control, and, most significantly in 1991, the U.S.-led effort to
liberate Kuwait in Operation Desert Storm. So even if Gorbachev could
not hold power for much longer, there was a good argument to be made
for backing him and working with his government while the opportunity
was there. Not surprisingly, then, some administration officials say
that they were indeed aware of the intelligence reporting that
indicated Gorbachev was likely to fall, and that their response was
to get as much as possible for the United States "before the window
closed." These objectives included completion of strategic and
conventional arms agreements, agreement over reunification of
Germany, and dissolution of the Warsaw Pact. They believed that if
they could reach such agreements with Gorbachev, they would be
"locked in" and accepted even if a harder-line government replaced
him.

Second, administration officials did not believe that they had anyone
better to support, even if it seemed that Gorbachev was doomed.
Yeltsin had made an unfavorable impression on administration
officials during his September 1989 visit to the United States. They
did not believe that he was committed to democratization, and
considered him personally unstable.

Moreover, administration officials did not believe that they had an
effective means for supporting anyone other than Gorbachev even if
they had chosen to do so. As Scowcroft put it, "Some of the analysts
kept saying that we should 'move towards Yeltsin,' but I didn't know
what that meant." Officials were also concerned that the very process
of seeking alternatives to Gorbachev would have the effect of
undermining him, possibly hastening his fall and leading him to
become less cooperative. Robert Blackwell, the National Intelligence
Officer responsible for the Soviet Union at the time, recalls that
the feeling in the administration was, "if you undermine Gorbachev,
you won't get Yeltsin; you'll get a successful coup."

Third, U.S. officials believed that they could influence events. That
is, the intelligence may have said that, as things stood, Gorbachev
would fall, but U.S. leaders were in a position to influence the
conditions on which these assessments were based. The intelligence
community's estimate was based on its evaluation of nationalistic
sentiment and the willingness of hardliners to take risks. It was at
least plausible that these attitudes could be shaped or moderated.

For example, Bush has been criticized for the speech he delivered in
Ukraine in early August 1991, warning of the dangers of nationalism
(the so-called "Chicken Kiev" speech). Some critics assert that it
started U.S. relations with the independent Ukrainian state on a sour
note; some argue that acquiescing in the dissolution of the Soviet
Union was actually in the interests of the United States. Whatever
one thinks of the policy, had the speech dissuaded the Ukrainians
from total succession, it would now be viewed as a classic example in
which the United States demonstrated leadership and influenced events.

The "Indications and Warnings"  Problem: Has the Coup Started Yet?

Scowcroft says unequivocally that he did not receive a warning that a coup was about to occur in mid-August 1991. He and other officials also note that they were in Moscow the week before the coup for the signing of the Strategic Arms Reduction Talks (start) agreement, and that not only were there no signs of the plot but many of the eventual plotters took part in the ceremonies.

According to Beschloss and Talbott, however, the President's Daily  Brief (PDB) for August 17 led with a report warning of signs of last-minute opposition to the Union Treaty. The treaty would have devolved most of the USSR's authority to the republics--exactly the development that previous CIA reporting had said might trigger a reaction by hardline Party, military, and KGB leaders. The PDB noted that the previous day Alexander Yakovlev had warned of an "influential Stalinist group" that was planning a "party and state coup."