Intelligent Reform

Journalists, lawmakers and President Bush gushed over the law he signed Dec. 17, 2004, to improve the performance of the 15 agencies that comprise the intelligence community. Veterans of that community were less impressed. "Pretty thin gruel," says James M. Simon Jr., former assistant director of Central Intelligence for administration. His assessment, informed by nearly 30 years at the CIA, is shared by many gray eminences of espionage.

The proponents of the 2004 Intelligence Reform and Terrorism Prevention Act, including its congressional authors, members of the commission that investigated the Sept. 11 attacks and family members of the victims, are especially proud that the law creates a director of national intelligence. They contend that the DNI finally will get intelligence agencies to act like a community.

Current and former practitioners of intelligence say the law's backers overstate the DNI's power. For instance, supporters contend the director will control the annual intelligence budget. Not precisely. The Defense secretary retains powers to craft budgets for Defense agencies and activities that comprise probably 80 percent of annual intelligence spending-exact amounts are classified. The law says the DNI will participate in those preparations.

Those who've seen the budgeting process up close believe a showdown is inevitable. The DNI's office and the Pentagon will arm-wrestle over which programs the DNI controls, says John Hamre, a former deputy secretary of Defense and now the president and CEO of the Center for Strategic and International Studies in Washington. "You can count on trench warfare over all the details," he adds.

Last year, lawmakers crafted reform proposals based on the 9/11 commission's findings during the height of the presidential campaign. Hamre cautioned against changes motivated by politics and Sept. 11. He and others implored lawmakers to enact the ideas of intelligence veterans who've grappled with reform in many guises for decades.

Many of their ideas didn't make it into the law and Hamre, Simon and others say the need for real reform remains. Their agenda includes improving the quality of intelligence analysis, addressing festering personnel issues, and considering whether and, if so, how the intelligence establishment should be reorganized.

Collecting vs. Analyzing

"Good organizations do one thing well," Simon says. And for intelligence agencies, that means collecting or analyzing data. But should they try to do both? No, reformers say. Each activity requires distinct skills, and often the two don't peacefully coexist. Historically, more money has flowed to collection, so the quality of analysis has suffered.

Separating collection and analysis might improve the performance of each, say Hamre and others. They would move the intelligence collectors out of agencies such as the National Security Agency, which intercepts electronic data, and the National Reconnaissance Office, which handles satellite imagery, and place them under the control of a single official. They would keep intelligence analyzers under the control of Cabinet secretaries and agency heads, as they are now.

The idea is to force organizations to focus on a single job and to make them prove their worth by how well they perform, explains John M. Poindexter, a former national security adviser to President Reagan. For several months, Poindexter has conferred informally with intelligence professionals to draft a vision of a realigned community that would rely on competition to elicit better results.

Poindexter proposes a director of national information to oversee collection operations-NSA, NRO, the National Geospatial-Intelligence Agency and the intelligence gathering activities of the CIA. "The collectors should have no other responsibility than providing their take to others," Poindexter says. "There would be no incentive to hoard information."

The analyzers, on the other hand, would compete with one another to provide the best analysis to policymakers. Because they'd have responsibility only for assessing information, not for collecting it, analyzers would gain clout purely by dint of superior divining. "Competitive analysis must be the expected standard," Poindexter says.

It's a vision shared by other intelligence experts. In a public proposal titled "Guiding Principles for Intelligence Reform," released in September, 11 former lawmakers and high-ranking security officials wrote: "The best analysis emerges from a competitive environment where different perspectives are welcomed and alternative hypotheses are encouraged. Intelligence reform must institutionalize these traits." The document was signed by, among others, former Defense Secretaries William Cohen and Frank Carlucci; former Senators Gary W. Hart, Bill Bradley and Sam Nunn; ex-CIA Director Robert M. Gates; and Henry A. Kissinger, who served as both secretary of State and national security adviser in the Nixon administration.

The role such decision-makers play in the intelligence process needs to change, too, reformers say. Under Poindexter's model, the National Security Council staff and the national security adviser-who stands between the president and the military, intelligence and diplomatic arms of government-would have more say in defining what should be analyzed, and therefore what should be collected.

Competition will breed contrarian thinking and multiple views on the same data. And Simon says that's a great thing. "Intelligence people need to fight over data and its interpretation," he says. "Being 'on the team' is antithetical to what their profession requires."

But that doesn't mean agencies shouldn't collaborate. Indeed, under the reform law, they must contribute to intelligence centers that the DNI establishes to study areas of interest. The law envisions a vast network of information technologies aiding analysis and providing ways to share data among agencies, as well as with state and local governments and the private sector. That will require strict security and privacy protection. But is the government prepared to implement such a system?

Three years ago, Poindexter studied analysis, information sharing and privacy technologies as part of the Total Information Awareness program at the Defense Advanced Research Projects Agency. He proposed culling private data for clues about terrorism, an approach that aroused controversy and helped kill the program.

The collaborative and privacy tools TIA studied were among some of the most innovative and aggressive the government ever has explored. Today they're not being pursued with the same vigor, even though government-sponsored and private groups have called on agencies to invest in data-sharing networks and tools that can distribute intelligence while protecting privacy. It's not clear whether the reform law can reinvigorate the research.

Current vs. Long-Term

Reorganizing agencies might position them to produce better analysis, but how can the craft of analysis itself be improved? Intelligence agencies, particularly the CIA, have struggled for more than 25 years with that problem. Analysis suffers from a lack of discipline and people's reluctance to study long-term problems, reformers say.

"Give me the information, and I'll tell you what it means," an analyst might say, instead of plugging known quantities of data into structures engineered to help solve particular problems. Such a structural methodology might produce a diagram that displays what an analyst knows and which pieces of information are connected.

Analysts need symbols, processes and rules to help them arrive at conclusions and defend positions, says a former CIA analyst who spent most of his career trying to instill methodologies. Analysis lacks a discipline such as the one lawyers follow when handling evidence, he says. He points to the work of American legal scholar John Henry Wigmore, who, in the early 20th century, created methods for organizing and managing evidence that have become standard practice for the legal profession.

Why is there no Wigmore for the intelligence business? It's not so much a disdain for methodology, says the analyst, who asked not to be named because he still works for the government. Rather, it's because analysts spend much of their time creating reports about current events.

This bias toward "current analysis," as opposed to longer-term studies, is another major deficiency, reformers believe. Current analysis is the intelligence version of daily news. It often finds its way into documents such as the president's daily briefing produced by the CIA. Current analysis often is produced at the expense of meatier reports that provide deeper insights, but might take months or years to complete.

Stansfield Turner, who headed the CIA from 1977 until 1981, attempted to nurture longer term, more comprehensive analysis. In 1977, the agency's analytic arm, the Directorate of Information, was renamed the National Foreign Assessments Center. It aimed for a more expansive, long-term view and utilized analysis from outside experts in addition to that of career intelligence employees. But the center was short-lived. In 1982, it reverted to the Directorate of Intelligence and current analysis remained the dominant focus.

Attempts to reform analysis continued. In the mid-1990s, a commission headed by then-Defense Secretary Harold Brown recommended restructuring the National Intelligence Council, the group that assembles national intelligence estimates based on the entire community's findings on a particular subject. The council would join a new National Assessments Center, an organization not unlike the CIA analysis center under Turner. There, the intelligence council would continue writing estimates, but "equally or more important, it should also prepare classified and unclassified 'assessments' of issues of concern to policymakers where the intelligence contribution is relatively small," the Brown commission wrote. "These assessments should include analyses of long-term problems policymakers are unable to address effectively because of their demanding schedules and the need to focus on current events." The proposal found its way into intelligence reform legislation, but Congress never passed it.

Generalists vs. Experts

With agencies reorganized and analysis addressed, reformers would home in on intelligence personnel problems. Some had hoped that the reform law would create an "intelligence service," establishing avenues for career enhancement and treating intelligence employees with more uniformity. Advocates of such a service point to the 1986 Defense Department Reorganization Act, known as the Goldwater-Nichols Act, which sought to simplify the military chain of command.

But intelligence reform didn't go that far. Instead, the new law simply states that promotion to certain positions designated by the DNI will require stints at more than one agency. Employees who serve in analytic as well as operational jobs get special consideration. The idea is to broaden the employees' experience, rather than to sharpen it on one issue.

And that troubles some, including Michael Scheuer, who led the CIA's hunt for Osama bin Laden in the late 1990s. Scheuer, who resigned in November, warns that intelligence agencies haven't cultivated terrorism experts. Instead, they've embraced generalists-often political scientists and historians-who sometimes write brilliant analysis, but lack the experience tracking terrorist networks that Scheuer believes is essential to defeating them.

"The intelligence community has contempt for expertise," Scheuer says, and historically hasn't regarded counterterrorism analysts as their brightest stars. Before the Sept. 11 attacks, Scheuer says, the CIA's Counterterrorist Center, which supports overseas spying and operations, was seen as a dumping ground for less talented analysts.

The intelligence agencies are full of experts. But Scheuer and others say there aren't enough of those who matter most now-terrorism analysts and linguists. What's more, they say, becoming a terrorism expert hasn't been the best way to enhance one's intelligence career. So, the problem isn't just a lack of experts, but a reluctance to develop and promote them.

CIA officials have said the agency is doubling its efforts to recruit and train a counterterrorism staff. They point out that the stable of officers assigned to the agency's Counterterrorist Center has expanded twofold since the Sept. 11 attacks. But Scheuer says this doubling consists of new employees who come to the CTC for an introductory lesson on terrorism. They stay, at most, for 90 days, and then move to other assignments, which may or may not be related to counterterrorism. A new crop replaces the trainees. Scheuer calls the staff increases a "shell game."

A July 2002 report by the House Intelligence Subcommittee on Terrorism and Homeland Security found that, from September 2001 to spring 2002, CTC employees "were not all experienced in the counterterrorism mission." The report urged the CIA to ensure "that home-basing for CTC case officers is a viable option and is career-enhancing."

Perhaps the CIA's personnel problems stem partly from the realities of modern employment. Scheuer says many recruits with whom he worked or met at the agency were young, bright, excited and prepared to work for the agency for about five years and move on. Young people don't envision working for one company all their careers. Why should they imagine they'd stay at one agency?

But the employment structure still is geared toward workers who stay in a job for two decades or more. Career advancement in intelligence, and government generally, is slower than in the private sector. Some intelligence rookies likely joined because they wanted to pursue a specific area of interest. But in a system that promotes slowly and favors generalists, what's the incentive for those people to stick around? Answering that question, reformers say, now falls to the DNI and the CIA director.

Meanwhile, President Bush has prescribed an ambitious personnel agenda for the CIA. Last November, he told Director Porter Goss to increase by 50 percent the agency's cadre of clandestine operators and analysts, as well as employees proficient in languages such as Arabic. That means the CIA will have to hire and train thousands of people over the next several years.

How many years is open to debate. Last April, George Tenet, Goss' predecessor, told the 9/11 commission that it would take the CIA five years to build a spy network that could effectively counter terrorists. Former CIA assistant director Simon says it could take a decade.

Management vs. 'the Fun Stuff'

The intelligence reformers, despite their dire assessments, aren't pessimistic. But they know how difficult it will be to move the entrenched intelligence bureaucracy. Since President Truman signed the law creating the CIA in 1947, more than 30 laws, commissions or proposals have recommended changes to the intelligence system.

Hamre says the future of reform depends on the DNI's willingness to manage and not be sidetracked by the sexier aspects of intelligence-espionage, intelligence collection, what he terms "the fun stuff." The DNI will have to become a super bureaucrat with credibility, skill and humility in equal measure, Hamre believes.

The enactment of a law won't silence the debate over how to repair intelligence agencies. But reformers agree on one point. If the new DNI isn't seen as a serious manager-and they likely will help make that determination-then the momentum for change will pass, at least until another intelligence catastrophe reignites the clamor for reform.