At his confirmation hearing, William Casey, the new director of central intelligence, stated that his primary objective as head of the CIA would be to improve the quality of the intelligence product.

There are two possible approaches to that task. Which one Casey and the Reagan administration choose will determine whether the intelligence community continues to be mired in constroversy.

One approach is to look backward and seek to undo the modest reforms relating to surveillance of Americans and Freedom of Information under the slogan of "unleashing the CIA." The alternative is to move the debate to a different level by focusing on proposals directly aimed at improving the intelligence product.

The former approach will do little to affect the quality of the intelligence that the president needs and much to continue the debate that has contributed to the declining morale of the intelligence agencies.

Despite all the rhetoric about shackling the intelligence agencies, they are in fact under very few restraints: most of the limitations relative only to the surveillance o of American citizens. The most restrictive limitations are not in executive orders or legislation but in agency implementing directives drafted by the agencies and approved by the attorney general. In urging the new administration to leave these directives in place, House Intelligence Committee Chairman Edward Boland (D-Mass.) noted that the current system has the support of the head of every intelligence agency.

Moreover, all of the post-Watergate restrictions taken together have only a very small impact, if any on the gathering of intelligence about the Soviet Union and other high priority targets; removing all of them will not improve the product in any significant way. Indeed, permitting the intelligence agencies to spy again on dissenting Americans could lead to a focus, as in the past, away from real counterintelligence efforts aimed at countering the KGB to the far easier business of surveiling lawful political activity.

On the other hand, if the new administration decides to leave the rights-of-Americans issues where it found them, neither seeking to undo the existing restrictions nor moving toward a legislated charter, it will, one suspects, find those who have been pressing for more reform willing to give the present system time to prove its worth.

This will leave the new team free to concentrate on the vital task of improving the intelligence product. Here they have an agenda, laid out among others by Richard Allen, President Reagan's national security adviser, which in view holds the promise of accomplishing the objective in ways fully consistent with the rights of Americans.

The key elements of this reform effort are: 1) the separation of the clandestine service for covert collection and operations from the rest of the CIA, which would become an analysis agency, 2) the encouragement of multiple centers of analysis and of competing estimates rather than joint intelligence community products and 3) the creation of a new intelligence coordinating position in the White House.

Each of these proposals would meet stiff resistance from some parts of the intelligence community; and Casey expressed general opposition to reorganization at his confirmation hearings, suggesting that there have been too many reorganizations. However, the fact is that these fundamental restructuring proposals have yet to be seriously considered. Taken together, they could significantly improve the quality of intelligence reaching the president.

The splitting of the CIA would permit the appointment of an analyst to the post of head of the CIA for the first time. (I am assuming that Casey would assume the White House intelligence role if these schemes were adopted.) It would permit that agency to concentrate on producing unbiased intelligence without responsibility for any collection programs or operations. One would hope that a tradition would develop of having the agency headed by a career official and a distinguished scholar from outside the government. The new agency could do much of its work without excessive secrecy and should be able to develop far more extensive and profitable relations with the research and academic communities.

This new intelligence agency should not devote substantial time to producing joint intelligence products. If such efforts are not proscribed, they should be limited to situations in which an agreed estimate is needed for planning purposes. Even then, every effort should be made to prevent the hiding of profound differences in carefully chosen, ambiguous language. Most of the product of the agency should be its own and should be signed by real people, not offices. Moreover, every other agency should be encouraged to develop and expand its own analytic capability and produce competing reports challenging the conclusions of the new analysis agency. The most compelling analysis should triumph, not the least common denominator of agreed estimates.

None of this will ensure good intelligence, let alone good policy. But it would start the intelligence community back on the road toward doing the job it was set up to do. It would also avoid the scrimonious public debate that can only prolong the period of decline in the quality of the intelligence product.