Earlier this year I resigned as National Intelligence Officer for Latin America because of the pressure put on me by the Director of Central Intelligence to come up with a National Intelligence Estimate on Mexico that would satisfy him. This is not the first time that pressure has been put on intelligence officers to come up with what their superiors consider to be the right answers. A previous director not long ago remarked that he was considered a "traitor" because the estimates on Southeast Asia that were being written under his direction were not pleasing to the policy-makers at the time -- the estimates didn't say that our policy in Vietnam was working. In my own case, it was not that the policy-makers were putting pressure on the director, but rather that the pressure on me and others working on the Mexico estimate came from the director himself.

Nothing will get an intelligence officer's back up faster than a sniff of that kind of pressure in his nostrils. It is a matter of principle that he not slant intelligence judgments to make them more palatable to his superiors or to shower the glory of approval on an administration's policies. A National Intelligence Estimate is not simply an intelligence report or a bit of analysis, nor should it be any one man's opinion. It is the product of the deliberation of representatives of all the intelligence agencies dealing with foreign affairs. As a member of the National Intelligence Council, the national intelligence officer chairs the writing of the estimate. Being in the chair may give him more influence than one of the representatives from CIA, from State or Army or Navy or Air Force or the Marines, or from the Defense Intelligence Agency. It may not. But the result should reflect the views of all the agencies and differences in their views. It is not or should not be blandly unanimous, and it should reflect doubts as well as disagreements.

In 1976m a distinguished intelligence officer, in testifying before the Senate, spoke of the "natural tension" between intelligence officers and policy-makers and said, "Policy-makers must assume the integrity of the intelligence provided and avoid attempts to get materials suited to their tastes." Much has been said -- and no doubt much more will be said -- about the motives of policy-makers for disputing or disliking the intelligence they receive. The point to understand and to accept is that this has happened in the past, and it can be expected in the future.

Strong-minded officials -- Republicans, Demcrats, career people of no partisan bias -- often think they know better than intelligence officers. Sometimes they don't care what intelligence says as long as it doesn't get in their way. Attempts to squelch displeasing intelligence reports or judgments that don't back up an administration's policies have a nonpartisan provenance. William Casey, the current director, most differs from previous directors of Central Intelligence in that he is a part of the policy-making group where Central America is involved as much as he is the president's chief intelligence officer.

His particular case has led to talk of a bill to ensure the selection of future directors from the career services to prevent politicians' being put in the job. That may appeal to us intelligence officers who have an unhealthy respect for our own virtue, but no legislation can ensure that a director, no matter how experienced in our work, will not buckle under pressure.

Ambition or the desire to go along with the gang -- to be on the team -- can lead us to ignore the warnings of conscience or of colleagues. Proposals for dealing with this problem discussed on a moral plane usually dissolve in empty righteousness. Legislation inoculates us against the disease from which we have just recovered without coping with the next set of symptoms. We should face the expectation that even men of good will and integrity my be intolerant of opinions they consider to be wrong or inconvenient. A taste of power may make us arrogant. The natural tension will continue.

If we accept this as inevitable, our aim should be to soften the collision. I propose that we do so through a loose, informal council of elders -- a tribal council -- to act as the public conscience, since intelligence matters cannot by their nature be thrown open to public scrutiny and since the early discussion of policy does not benefit from speech-making. The council would sit with the director when he is beleaguered by the politicians, hold his hand when temptation beckons him from the path of duty, and talk quietly with other parties to see if the differences be minor or major and to sound warnings if the risks to be run seem not worth the candle. The council would be made up of members of the four different organizations already charged with the task of examining the performance of the intelligence community and of the CIA in particular.

In the CIA there is an Office of the Inspector General that inspects the agency and acts as ombudsman for employee complaints. The President's Foreign Intelligence Advisory Board is made up of private citizens appointed by the president. Two other organizations charged with oversight of the intelligence community are the Senate and House intelligence committees. The informal exchange of information and views among these groups would provide an immense improvement.

What would begin as a pragmatic approach to supporting the integrity of the intelligence process could benefit from the participation of officials from State, from Defense. The discussion of other differences in foreign affairs in discreet, informal settings could accomplish more than the noisy and grudge-making spats that too often accompany public arguments. The rhetorical sharpness of incoming administrations would sooner be honed by the starkness of the confrontation with real problems and their obdurate nature. The capture of policy strong points by wrongheaded little ideological factions would be less likely.

Good intelligence is vital to our security. Our discussion of foreign and defense policy suffers grievously from partisan exaggerations and simplifications. A tribal council, talking over intelligence judgments, could build bridges over petty chasms, define real differences and increase the area of consensus that seems so far from our grasp today.