True intelligence reform is cultural, not just organizational chart shift

The intelligence reform bill recently signed by President Bush will hardly solve all the problems confronting the American intelligence community, but it is a beginning. Someone is at last in charge. The new director of national intelligence (DNI) will not, though, have a free hand to reshape American intelligence. The bill gives the DNI authority to move only small amounts of money. So big changes - like collapsing all the intelligence collectors into a single agency - are ruled out.

Creating a DNI always was going to be just a first step. In pushing that measure, the 9/11 commission did an impressive job of selling a good idea that had almost nothing to do with 9/11. The failures that the commission documented are operational - too little sharing of information and insufficient attention to the counterterrorism mission, especially by the FBI. An intelligence director is only indirectly related to those failings.

In the short run, the simple fact of 9/11 has impelled better cooperation among the different elements of the intelligence community. Sure, in the long run, having someone in charge should make for better cooperation between the FBI and the CIA. But as two generations of secretaries of Defense have found in pressing for more "jointness" among the military services, the task is long and arduous. Real reforms are much more matters of organizational culture than of organization charts.

Improving analysis is one of the most important initiatives for the next phase. The 9/11 commission was eloquent about the need for more creativity, but its recommendations do not really touch that topic. Indeed, its main recommendation in the final bill - to organize the analytic part of the intelligence community around issues, as with the National Counterterrorism Center - has merit but also carries risks. The centers will be consumed by the need to provide the very hottest current intelligence.

The real challenge is to develop a cadre of intelligence analysts who are encouraged to "think creatively" and to acquire intellectual capital in the form of substantive expertise on a broad range of topics. Threats to the US, like terrorism, are global and adaptive, blurring distinctions between crime, terrorism, and war. Analysts of the future need to think more like homicide detectives who focus on solving puzzles with incomplete information.

The need for creative thinking runs directly into the need to reform secrecy and compartmentalization. Current rhetoric about the "transformation" of the intelligence community celebrates exploiting information from the full spectrum of secret and open courses. Such terms as "multi-intelligence" and "fusion analysis" are the catchphrases. The favorite is "connecting the dots." Unfortunately, the current system is biased toward compartmentalizing information, not sharing it. To protect the US, every agency controls its own information, with access granted on a "need to know" basis. Yet creativity in analysis will come precisely from having people with no "need to know" look at data, because they may see patterns that current experts do not.

Workforce reforms are another important challenge. All the intelligence agencies have grown substantially since 9/11. The growth is a wonderful opportunity. The new young intelligence analysts are fearless and computer savvy. They will not stand for the information environments - compartmentalized, slow, and source-driven - that current intelligence provides. Yet they are also untrained, and given the aging of the intelligence agencies, will lack for mentors. All the agencies will have to dramatically rethink how they deal with the lives and careers of their premier asset - their people. At present, training of analysts, in particular, is mostly on the job. And almost all training is stovepiped by agency, so analysts have little idea how counterparts in other agencies work.

Yet another challenge is finding a way to keep the intelligence community from drowning in information. For one thing, while terrorists are secretive, data that is not secret - phone numbers, drivers' licenses, and the like - is also relevant. For another, America's technical capabilities to produce secret information, such as imagery from spy satellites, have mushroomed. In particular, the big collectors of imagery and signals intelligence - the National Geospatial Intelligence Agency and the National Security Agency - will be tempted to turn a firehose of data on intelligence analysts. After all, that is the way to make sure that they are not the culprits for the next intelligence failure.

A better balance is needed between investments in the emerging collection systems and enhanced forms of analytical capability. The latter means a greatly expanded investment in quality personnel and new technologies that assist analysts, instead of overwhelming them. Put simply, huge amounts of data collected but unprocessed and unanalyzed are useless to the policymaker.

Finally, there is a need for new forms of intelligence collection. Every blue-ribbon panel calls for improving America's espionage, or human intelligence capabilities. The call is worthy, but expectations have to be reasonable. However one judges the past half century of US espionage, doing better against tomorrow's much harder targets, like terrorists, will not be easy. The required actions - making much more use of America's ethnic diversity, or moving spying out of official US government buildings, for instance - take time and money.

The intelligence reform bill should be viewed as the necessary first step, but hardly as sufficient. This next phase will require leaders in the intelligence, national security, and law enforcement communities willing to take risks. Most important, Congress needs to be convinced that what it has done so far is just the beginning. The people of the US intelligence community are up to the task of real reform. The question is whether their current and prospective leaders are also up to that task.

• Gregory Treverton is director of the Intelligence Policy Center at the nonprofit RAND Corporation. Peter Wilson is a senior analyst at RAND.