11 January 2008

Intelligence Oversight

Balkin's blog notes the current problems with oversight of intelligence agencies. This has been an open secret for a long time.

The 9-11 commission got it fundmentally wrong when they urged a national intelligence czar. Connecting the dots isn't something that happens at the senior management level. It requires massive sharing of information, so that people who know something share information that they don't realize is important to people who need to know something.

In my view, the key step is to bifurcate the intelligence community, so that the bulk of the intelligence community, the analysts and spy satellite operators and the like, who have less opportunity to engage in egregious conduct and deal with a large volume of material that isn't "life or death secret," from the smaller group of human intelligence operatives who are in a position to misbehave and also have much more sensitive methods and practices information.

The former should operate in a much more open, less secrecy obsessed environment (to the extent possible in the same agency as the diplomats who are their primary customers) and should be permitted to operate with less direct oversight from the political branches. This agency might have 90% of the employees and budget authority (the details are secret, these are just illustrative numbers) and 10% of the oversight effort. The latter can't afford to have such an internally open agency culture, but needs to compensate for that by having far more intense political supervision because their activities are much more susceptible to abuse. One can do some very good big picture intelligence without relying on covert sources whose identities and methods are highly sensitive. Huge CIA failures like its inability to see the impending collapse of the Soviet Union didn't have anything to do with a lack of undercover agents able to describe the contents of senior officials' desk drawers and hard drives.

The smaller, more covert, part of the operation is in a poor position to share intelligence within the agency because methods and practices concerns are so vital. So, this agency ought to have a very flat hierarchy. Regional or country managers should be Presidentially appointed and Senatorially approved officials who report findings directly to the White House office and Congress, with a thin head office handling H.R. and inevitable agency management functions, while mostly insuring the orderly succession of regional agency managers. By and large, the head of this smaller agency should not be in charge of filtering information flow aimed at policy makers, a job better sued to a senior White House office official. It is much harder to dodge responsibility for misconduct and poor results in a very small bureacracy than in a large one. Also, this kind of structure would reduce the likelihood of an inevitable bad apple in one component of the agency tainting another part of the agency.

No comments: