Too Juicy To Keep Secret

IF JEAN-BERTRAND ARISTIDE IS mentally unfit to rule, why has the U.S. government built its Haitian policy around him? If he isn't, why does the CIA keep telling the Congress that he is? There may be good answers to these questions, but so far we have not heard them. What we have heard is the sound of pols and spooks going bump in the press--yet another collision of intelligence with policy.

The current uproar began when a senator began disclosing the results of a CIA profile questioning Aristide's fitness to govern. The administration dismissed the CIA report, claimed that its own experience with Aristide proved the contrary and reminded everyone that he was elected president in 1991 with two thirds of the vote.

The Aristide case is the latest reminder of how unquiet the relationship between intelligence and policy can be. That's especially true in the United States, where our intelligence community is uniquely susceptible to politicization, in part because it answers to many masters. In countries like Britain, the intelligence business is watched over by a clubby coterie of ministers and senior officials who guard its system and secrets and keep public ventilation out. That's how it generally was in the United States until congressional hearings during the 1970s drew attention to abuses. Ever since, wider oversight has both benefited and burdened the American system.

Since our system isn't likely to change, we need to deal with its consequences. There may be one Director of Central Intelligence, but the intelligence community must respond to a panoply of bosses within the executive branch, as well as competing intelligence agencies and the House and Senate committees that monitor--and fund--intelligence activities. Any and all of the players can turn the disclosure of intelligence analysis into a political act. In Washington, if anybody can, somebody will.

No one's hands have been entirely clean. Many intelligence officials are well aware that their congressional briefings can complicate--and undermine--policy. Some senior ones often have strong policy views, which may conflict with those of their bosses in the executive branch. Invariably, bosses in every administration decry intelligence officials for "spilling their guts on the Hill and undermining our policy."

These political dramas draw their material from the complicated process of intelligence gathering and analysis. The quality of intelligence analysis varies greatly. For example, even after 40 years of targeting, North Korea remains largely unknown to us. Haiti, on the other hand, should be an open book, because the country is accessible and so many Haitians reportedly have been on our payroll. That hasn't been enough to narrow the differences of view. Whether the CIA's involvement with anti-Aristide factions has distorted reporting and analysis is hard for an outsider to judge.

Intelligence analysis is the weakest link of the intelligence process. Many analysts have little firsthand familiarity with the subjects they write about. They tend to trust only information that is bought, overheard or photographed. Even now, after Aristide's long residence in the United States, does the CIA's psychological profile still stand up? How up to date is it? The CIA isn't likely to repudiate its original report--if for no other reason than were it to do so now, it would be accused of caving in to political pressure.

Intelligence analysis is not a search for metaphysical truths. it is mostly a painstaking and highly bureaucratic effort to help policymakers reduce uncertainty--to help them decide whether or not Aristide is mentally fit to govern, not whether he is the "moral equivalent of George Washington." On many subjects the uncertainties are so great that intelligence analysis can be used publicly by competing groups to support or attack current lines of policy at one and the same time. That's precisely the point at which an administration must make a sound policy judgment--and a solid public case--for a course of action.

Despite their avowal to the contrary, policymakers tend to choose the analysis that supports their policy and discard the others. In the Aristide case, the administration either chose to ignore the analysis based on its own experience (Aristide was apparently impressive in his meetings with President Clinton and others) or because it decided it had to support Aristide since he was the last and very popularly elected president. The administration hasn't said much about the question of Aristide's reluctance to condemn violence on the part of his own supporters. It also seems to discount the view that Aristide is likely to be killed when he returns. It would appear that the administration did not early on acknowledge its intelligence dilemma to Congress and make the case for its political decision to support Aristide nonetheless.

Tensions between intelligence and policy will always arise in our system. Legislators and bureaucrats will go on leaking for their own purposes. The best approach is simply to focus on improving the intelligence and the policy. That way, when leaks occur, spooks and pols alike will end up standing on higher ground.