William E. Odom

No organizational design will compensate for incompetent incumbents, but some designs prevent competent incumbents from performing well. The 9/11 commission's design for a new national intelligence director (NID) is sure to accomplish the latter. There is already a layer of bureaucracy above the CIA, NSA, DIA and other intelligence agencies, and it consists of the Community Management Staff and the National Intelligence Council. It simply has not been used effectively because the director of central intelligence is double-hatted -- that is, he is both CIA director and coordinator of the nation's intelligence agencies. Creating a NID with three deputies -- for homeland security, domestic intelligence and foreign intelligence -- would make things much worse. It would assure turf battles and prevent effective budget management.

Several senators have endorsed a fixed term for the NID, saying this would prevent the post from being politicized. It's a bad idea. The FBI director's term is fixed, and more often than not, this arrangement has fostered bad relations between the FBI director and the president. It would create worse ones for a NID. There is no way to depoliticize the role of the president's intelligence chief. It is a desirable aspiration, but intelligence is just as political as policymaking and military operations.

The popular notion that apolitical intelligence will prevent bad policies is an illusion. Intelligence chiefs can be no more effective than their political leaders or military commanders allow them to be or demand that they be. The intelligence failures surrounding the 9/11 attacks and in Iraq are primarily political failures. Effective leaders do not tolerate inadequate intelligence performance or leave it to commissions to fix intelligence problems.

The following historical anecdote may be instructive:

In late 1944, as the German Wehrmacht prepared to launch its last counteroffensive at the Battle of the Bulge, several pieces of intelligence suggested it was coming. The top American generals couldn't agree on the value of the intelligence. Montgomery and his obedient intelligence officer (known as his G-2) stubbornly rejected the facts; Bradley and his G-2 remained skeptical and passive. Eisenhower and his G-2 were somewhat quicker to sense the danger but slower than Patton, whose G-2 saw it coming several weeks beforehand, prompting Patton to get his divisions ready to meet the offensive.

So four commanders with essentially the same intelligence turned in different performances. Though slightly disadvantaged by being at a lower echelon than all the others, Patton was far ahead of them in his appreciation of the impending assault.

In writing up this case study, Harold Deutsch, a military historian in World War II, showed how the personalities of these commanders intimidated their G-2s, discouraging them from emphasizing unpleasant findings or pursuing other lines of analysis. In his words, "Whether the commanding general was on the correct or wrong track, therefore, the G-2 was likely to be right there with him. Perhaps the fine performance of Gustave Koch [Patton's G-2] was largely due to being lucky in his boss."

When we ask how to improve the intelligence community's performance, we must recognize that it cannot be much better than the performance of the policymakers and commanders who own it.

William E. Odom, director of the National Security Agency from 1985 to 1988, is a senior fellow at the Hudson Institute.