Despite repeated criticism of weaknesses in CIA analyses of terrorist activities and Saddam Hussein's Iraq weapons programs, the new intelligence reform act requires several key CIA analysis practices to be enforced throughout the entire intelligence community.
The legislation, signed into law Friday by President Bush, asks that the new director of national intelligence designate people to enforce those practices, which have been in effect for years within the CIA.
"The CIA has been there and done that," said a former senior administration official who served in both the intelligence community and the White House and who reviewed several of these provisions applicable to intelligence analysis.
For example, the new national intelligence director must pick an "individual or entity" to be responsible for ensuring that "elements of the intelligence community conduct alternative analysis of the information and conclusions in intelligence products." Such analysis is commonly referred to as "red-team analysis."
The CIA has had red teams, or "red cells," doing critiques of its products or reports for years. "There is a daily red team now within the Directorate of Intelligence," the CIA unit that does analysis, an intelligence official said.
One problem with the technique, a former senior analyst said, is that such teams sometimes find gaps where none exists and thereby delay a timely distribution of an analysis paper while the original authors try to answer the criticisms.
Another CIA practice being spread to the entire community is to have a quality control office or officer make sure that analyses conform to high standards. The law requires the intelligence director to assign someone to ensure that "finished intelligence products" are "timely, objective, independent of political considerations, based upon all sources of available intelligence and employ standards of proper analytic tradecraft."
Because hundreds, if not thousands, of "finished intelligence products" are done by analysts in 15 agencies, this could be a gigantic undertaking, particularly because the law requires this oversight to be performed "on a regular basis" on one or more elements within the community "covering a particular topic or subject matter."
The law says such reviews should include the study of all available sources and those used, and the "quality and reliability of underlying sources." Analysts not being told about the reliability of clandestine sources has been seen as a longtime problem.
The reviewers should also see whether such sources carry proper caveats or "express uncertainties or confidence in annalistic judgments" and "properly distinguish between underlying intelligence and assumptions and judgments of analysts." Every year the national intelligence director must report to the congressional oversight committees on such investigations.
The CIA has had such a product review staff for some time. It was established, a former senior analyst said, to give the top officials a sense of the reports and of messages being sent to customers, such as policymakers and the administration.
The agency also has an ombudsman to whom analysts and others can raise concerns about problems that do not require a full investigation by the CIA inspector general. Complementing that position, the new law requires the national intelligence director to appoint an individual who would provide that function just within the director's office and what is termed the "safeguard of objectivity in intelligence analysis."
The person "would be available to analysts" just within the director's office to "counsel, conduct arbitration, offer recommendations, and, as appropriate, initiate inquiries into real or perceived problems of analytic tradecraft or politicization, biased reporting or lack of objectivity in intel analysis."
The CIA ombudsman was initiated in response to the era of CIA Director William J. Casey in the Reagan administration, a former analyst recently said, when orders appeared to come "from the top down" to keep revising reports in order to finally reach certain analytic conclusions.
