Pre- 9/11 Intelligence Problems Still Exist

Top intelligence analyst Thomas Fingar says the Intelligence Community is doing well under reforms. But according to a new study by the Rand Corporation, post-9-11 intelligence the reforms have not solved the problems that enabled al-Queda’s plot to go undiscovered.  

Thomas Fingar, the Deputy Director of National Intelligence for Analysis, or DDNI(A), reported to the Committee on House Armed Services this February:

“I am pleased to report that the Intelligence Community is even better than it was last year as a result of the continuing implementation of reforms required by the Intelligence Reform and Terrorism Prevention Act of 2004. [There is now] more innovative and rigorous analysis and wider and more far-reaching collaboration.”

However, according to the Washington Times’ Bill Gertz, Fingar’s praise of “supposed improvements in U.S. intelligence analysis” doesn’t hold up under RAND Corp.’s February report on the Intelligence Community’s performance. The report, “Assessing the Tradecraft of Intelligence Analysis,” states there are still concerns intelligence agencies haven’t moved past pre-9/11 problems in their analysis and integration.

Buried in the report, within a statement about the need for more long-term analysis, the authors said there is still inadequate judgment of sources going on in the intelligence community:

“[D]ata owners and processors [are] becoming confined to single sources, unable to take the broader view and to know what other sources might offer. … Analysts often know too little of the sources on which they depend, especially the human sources, to judge their reliability.”

This results in a false sense of security about the reliability of intelligence analysis and judgments which — in the compressed environment in which intelligence is gathered, analyzed, and used — often snowballs small misjudgments into huge errors. And those errors are then multiplied and transmitted to intelligence consumers.

Among the lingering pre-9/11 problems the Rand report addressed:

“[The Intelligence Community] will also need, and need to communicate, a more explicit understanding of the limits of analysis under varying conditions and circumstances. Consumers often want definitive answers — point predictions — but intelligence can rarely provide them and should seldom pretend to try” (10).

One illustration of this problem is the recent NIE on Iran co-authored by Fingar. The media had a field day — month, really — when the report stated with “high confidence” that Iran had halted its development of nuclear Weapons of Mass Destruction (WMD) in 2003 because of pressure from the international community.

The problem with reporting that information with “high confidence” is clear when reading some of Fingar’s most confusingly ambivalent statements to the House last month:

“We judge with high confidence that in fall 2003, Tehran halted its nuclear weapons design and weaponization activities, as well as its covert military uranium conversion and enrichment-related activities, for at least several years. Because of intelligence gaps, DOE and the NIC assess with only moderate confidence that all such activities were halted. We assess with moderate confidence that Tehran had not restarted these activities as of mid-2007, but since they comprised an unannounced secret effort which Iran attempted to hide, we do not know if these activities have been restarted.” (p. 6 of transcript, emphasis added).

The problems with that report have been described in detail by critics such as Rep. Pete Hoekstra (R-Mich.), ranking Republican on the House Permanent Select Committee on Intelligence.  Hoekstra criticized the report for bootstrapping questionably-sourced intelligence into the “high confidence” conclusion. 

That report has since effectively stopped international action against Iran for its nuclear pursuits. 

Another pre-9-11 problem the RAND report said still plagues US intelligence operations is the intelligence community’s lack of collaboration between agencies. The report said separate agencies still aren’t playing well with others:

“Analysts from one agency … are mostly ignorant of one another. The need for a focal point in analysis and analytic tradecraft is striking, and this need will only grow as the Community strives to be more ‘joint’ in the wake of the December 2004 intelligence reform law and the creation of a director of national intelligence.”

The authors suggested “a more horizontally distributed Intelligence Community,” with different “cells” focusing on specific problems, with different agencies being represented within each of the cells. They also suggested more use of outside expertise, more data-sharing between agencies and a Community-wide means of training the next generation of analysts.

The Intelligence Community needs to be forced into a state of “jointness” before Fingar can rightfully claim there is “more innovative and rigorous analysis and wider and more far-reaching collaboration” going on.

Not enough has changed. The December 2004 reforms that created Fingar’s job, among others, have not solved the problem. Separate agencies still play separately. Sources are still limited. And the Intelligence Community appears to be in a state of denial about its problems.