The fabled CIA operative James Jesus Angleton once described counterintelligence as a wilderness of mirrors. Nothing is quite what we see; it may be a reflection of another reflection. The blue-ribbon panel that will review U.S. intelligence on Iraqi weapons of mass destruction will surely dig deeply into intelligence's dealings with Bush administration officials, but it is worth pausing over the intelligence side, how we knew what we knew, or didn't know. In a perfect storm of inference, ignorance and (dis)information, we convinced ourselves that Saddam Hussein must still have WMD.
The critical intelligence assessment, the October 2002 National Intelligence Estimate, was pre-eminently an inference from Hussein's previous behavior. What we knew is that he had had WMD programs and stockpiles of, especially, chemical weapons. Hussein had a long history of brandishing those programs. As a result, almost all intelligence analysts shared a strong preconception with policy officials: Hussein must still have WMD programs.
Ignorance reinforced that preconception. The review panel will look harder, but from what we now know, there seems to have been surprisingly little new information collected after the U.N. inspectors were expelled in 1998. If this was an intelligence failure, the failure lies there, with collection. Lacking a diplomatic presence in Iraq, the United States seems not to have had much HUMINT, or human intelligence - that is, spies close to Hussein. For HUMINT, it was dependent on friendly intelligence services, and so found it harder to calibrate what it received. British HUMINT sources in Iraq are described in some detail in the recent report chaired by Lord Hutton.
The United States did have technical assets, both signals intelligence, or SIGINT, intercepted Iraqi communications and other signals; and imagery, pictures or other images taken of Iraqi facilities. But there were two problems with those sources. Imagery can identify big nuclear facilities but not necessarily small stockpiles, especially if they are hidden. And much of U. S. signal intelligence apparently was devoted to intercepting tactical Iraqi military communications to provide warning of any threats to U.S. and allied pilots patrolling the northern and southern no-fly zones in Iraq.
Our ignorance also may have given dubious information more of a hearing than it deserved. Iraqi exiles -- Ahmed Chalabi prominent among them -- long had been hawking tales around Washington, and in the process generally discredited themselves in the eyes of U.S. intelligence. The exiles and their stories got a renewed hearing, through Vice President Dick Cheney's office and the Pentagon's newly created Office of Special Plans, the latter poring anew over the WMD case to see if the CIA and other analysts had missed something.
Finally, there was Hussein's curious disinformation campaign. Even after he readmitted the U.N. inspectors, he still behaved like a man who had something to hide. By not taking the steps to prove he didn't have what he didn't have, he reinforced the impression that he must have something to hide.
Why he behaved that way remains a puzzle. Perhaps, in some combination of his own befuddlement and the reluctance of his lieutenants to level with him, he actually believed he still had stockpiles. Or perhaps simple pride before the Arab world induced him not to admit he had relinquished the WMD card. Or perhaps he calculated, wrongly, that the Bush administration would react the same as the Clinton administration had, that WMD had prevented Clinton from attacking Iraq and would do the same in Bush's case.
In the end, most analysts inside and outside government -- including those who opposed going to war when the United States did -- would have bet that Iraq still had WMD programs and some stockpiles, at least of chemical weapons. Whether Iraq had WMD was a puzzle, not a mystery; there was a definitive answer in principle. Yet, as with most important intelligence issues, the analysis began where the evidence ended.
The United States got some negative evidence once the U.N. inspectors returned to Iraq. But negative evidence could not provide a convincing negative answer to the question: Does Iraq have WMD? At least it could not provide a "no" answer given strong preconceptions that the answer was positive -- not to mention the pressure of an administration that was, as President Bush recently has said, determined to make war on Hussein as a last resort for many reasons, not just WMD.
The failure goes deeply into how we think we know what we know, and so is not one to be remedied with more money or more human intelligence or other suggestions that will come first to the mind of the review panel. The failure is a reminder that intelligence is inference; it is not evidence. Nor is it policy, and the failure is also a reminder of the danger of turning a policy question -- should the United States eliminate Saddam Hussein -- into an intelligence issue, or appearing to do so.
Gregory F. Treverton is a senior analyst at Rand Corp. and associate dean of the Pardee Rand Graduate School. He was vice chair of the National Intelligence Council in the first Clinton administration, and his book "Reshaping National Intelligence for an Age of Information" was published in 2001 by Cambridge University Press.
This commentary originally appeared in San Francisco Chronicle on February 15, 2004. Commentary gives RAND researchers a platform to convey insights based on their professional expertise and often on their peer-reviewed research and analysis.