The Central Intelligence Agency is trying to convince itself and the American public that it provided U.S. decision makers with timely warnings about the impending collapse of the Soviet Union. The latest revisionist effort came last month, when the agency co-sponsored a three-day conference at Texas A&M's George Bush School of Government and Public Service on "U.S. Intelligence and the End of the Cold War." Simultaneously, the agency released a 378-page volume of 24 newly declassified documents that, officials claimed, showed the CIA did far better than its critics have said in anticipating the Soviet Union's demise in 1991. "The assertions that the CIA got it blatantly wrong are unfounded," was how one agency official put it.

At best, this is an exercise in self-deception. As one of the agency's Soviet analysts during the 1980s, I had a front-row seat from which to view our performance; if I had been invited to give my assessment at the Texas A&M conference, I would have offered a far more dismal--but, I believe, far more accurate--evaluation of how we did.

First, it is important to define the terms of this debate. The agency is engaging in a sleight of hand in making public a set of documents produced from 1988 to 1991. The CIA cannot be applauded because it warned in 1989 that Gorbachev's prospects were "doubtful at best" or, as one of the declassified document shows, because it said in November 1990 that "the Soviet Union as we have known it is finished." By that time, the endgame was upon us and most discerning observers could see that the Soviet regime was in trouble.

Instead, the agency should evaluate itself by looking at the entire record from the 1980s, particularly after Mikhail Gorbachev came to power in 1985. If it were being honest (and some speakers at the Texas A&M conference made some of these points), the CIA would say that it had exaggerated the strength of the Soviet military and economy, underestimated the burden of Soviet defense spending and ignored Gorbachev's efforts to engage the United States in a series of disarmament agreements.

The full record shows that the CIA did not anticipate Moscow's retreat abroad and its vulnerability at home. During the mid-1980s, the agency issued only limited warnings of Soviet weakness and nothing to indicate that the strategic relationship between the United States and the Soviet Union was about to change radically. As a result, it could not predict, anticipate or even speculate about the consequences of the Soviet decline. The costs of failing to achieve these most basic goals of intelligence work are tangible: the huge defense budgets of the Reagan era; a needlessly prolonged confrontation with Moscow; and a lost opportunity to influence developments in the Russian Federation.

I'm not alone in my view of the agency's failings. Various Reagan administration officials--including George Shultz, who was secretary of state, and Colin Powell, Reagan's last national security adviser--have said that they were not well served by the CIA's estimates (as these forecasts are known in the intelligence community). Part of the problem was political: William Casey, Reagan's director of central intelligence, only wanted information that would support his view of the Soviet threat and a strategy of confrontation with Moscow.

There's an argument to be made (and some have done so) that the U.S. military buildup helped put pressure on the Soviet Union, but that's not a defense for faulty intelligence work. Such a strategy is the province of policy makers; the CIA's job is to provide its best analysis of the facts.

Controversy over the CIA's performance broke out as soon as the Soviet Union began disintegrating. When the Senate Select Committee on Intelligence held hearings on the issue in 1991, I was asked to testify. I described the politicization of intelligence that occurred during Casey's tenure, including estimates that were skewed to undermine Shultz's efforts to improve relations with Moscow.

In 1985, Casey and his deputy director for intelligence, Robert Gates, had orchestrated an assessment of a nonexistent Soviet plot against Pope John Paul II. And a wildly off-the-mark estimate of a Soviet threat to Iran had been used to justify arms sales to so-called moderates in Tehran. Gates later denied that he had any role in politicizing intelligence, but said that he watched Casey "on issue after issue sit in meetings and present intelligence framed in terms of the policy he wanted pursued." This is a remarkable admission from Casey's primary intelligence adviser from 1981 to 1986.

In my testimony, I described how Casey and Gates turned aside numerous studies that my colleagues and I drafted; these studies documented Gorbachev's efforts to restructure the Soviet system and to pull back from its involvement with the Third World. I drafted a memorandum in 1985 concluding that the Soviet Union would not provide MiG-29s to the Sandinista government in Nicaragua; Gates rejected the memorandum and stated that the CIA should not "go out on a limb" on this issue. This was Gates's standard formulation for assessments that did not coincide with Casey's perception of the Soviet threat.

During my 24 years at the CIA, I saw occasional efforts to skew analytical conclusions. But these forays always seem to come from outside the agency. During the Vietnam War, for example, there was pressure from Johnson administration officials to understate the strength of Viet Cong forces. In the 1980s, however, Casey and Gates worked from within to slant intelligence on numerous policy-sensitive issues, particularly Soviet issues.

Let me be clear: My intelligence memorandums during this period--and those of my colleagues--did not predict the collapse of the Gorbachev regime, let alone the end of the Soviet Union. But if our unedited assessments had been released to policy makers, I am convinced the Reagan and Bush administrations would have had a better and earlier understanding of the changes underway in the Soviet Union.

For example, CIA economists had tracked the early stages of decline of the Soviet economy from 1976 to 1986, but Gates would not circulate most draft assessments that pointed to Soviet weakness. As a result, CIA estimates overstated the size of the Soviet economy and underestimated the economic burden of maintaining the Soviet military. During this period, the CIA estimated the size of the Soviet economy to be nearly 60 percent that of the U.S. economy and asserted that the growth rate of personal consumption in the Soviet Union from 1951 to 1988 exceeded that of the United States. (As late as 1986, the agency was asserting that East Germany was ahead of West Germany in per capita output.) The Reagan administration used these exaggerated numbers to justify nearly $2 trillion in total defense spending in the 1980s.

Part of the CIA's problem was a result of the fact that the agency was dead wrong on the most crucial Soviet intelligence question of the 1980s: Was Gorbachev serious? The CIA missed virtually every sign of change during the Gorbachev era, beginning with the significance of his accession to power, the political impact of the appointment of Eduard Shevardnadze as foreign minister, and the revolution in disarmament policy. CIA estimates provided no early warning of the Soviet retreat from the Third World, including withdrawal of its forces from Cam Ranh Bay in Vietnam, the Indian Ocean and the Mediterranean.

The CIA too quickly dismissed Moscow's claims that it would withdraw from Afghanistan--the first major step in Gorbachev's strategic retreat. His Afghan withdrawal foreshadowed his decision to not intervene at the fall of the Berlin Wall, setting the stage for anticommunist revolutions in Eastern Europe and the reunification of Germany. The CIA's misreading of Gorbachev meant that the agency could not anticipate these events, thus missing the greatest triumph of political liberalism in modern history.

The agency also distorted the military might of the Warsaw Pact and never anticipated that the pact would dissolve. These distortions delayed negotiations on a Conventional Forces in Europe Treaty. In 1990, only months before the pact's collapse, the CIA concluded that the Warsaw Pact had matched or exceeded NATO's capabilities in all ground-force weapons and would keep pace with NATO's modernization program.

Even as late as 1988, one estimate had concluded that the "domestic problems of the USSR are unlikely to alter the Soviet system and the international appetites that spring from it." And 1988 is part of the period when, agency officials insisted last month, the CIA got it right.

Before the CIA can expect to convince the public that it got it right, the agency will have to convince those policy makers who said they were surprised by the disintegration of the Warsaw Pact and the Soviet Union. Former president George Bush and his national security adviser, Brent Scowcroft, recall no CIA warnings about the Soviet demise or the collapse of the Berlin Wall. Powell has written that the CIA could not "anticipate events much better than a layman watching television." Even Gates conceded in his memoirs that the agency had underestimated the dramatic change of course in Soviet policy and had neither anticipated Gorbachev's retreat abroad nor the destruction of the Soviet system at home.

But the most compelling voice is that of former secretary of state Shultz, who believed that CIA analysis on the Soviet Union was "distorted by strong views about policy" that blocked any discussion of Soviet weakness. In 1986 and 1987, according to his memoirs, Shultz confronted CIA director Casey and accused him of providing "bum dope" to the president and warned the White House that the agency was "unable to perceive that change was coming in the Soviet Union." And in 1987, he reminded acting director Gates that the CIA was "usually wrong" about Moscow and that the agency had dismissed Gorbachev's policies as "just another attempt to deceive us."

The key to a genuine post-mortem is to examine all points of view on the subject. At last month's conference, however, the CIA did not want to hear from critics of the agency's track record, including me and two other former CIA officials who testified against Gates's appointment in 1991 as director of central intelligence. I asked the CIA's organizer of the conference to accommodate one of us, but he dismissed the idea.

If the CIA merely tries to exonerate itself, it cannot learn from its mistakes. That has happened too often in the agency's past. It should not happen now.

Melvin Goodman is a professor of international security at the National War College and a senior fellow at the Center for International Policy. He was a Soviet analyst at the CIA from 1966 to 1990.