Tenet on September 11, 2002, the day before Bush’s speech to the United Nations.
Tenet refused the request on the grounds that Graham wanted an assessment of U.S. strategy and policy.
That was way outside Tenet’s purview. The CIA made assessments and formal National Intelligence Estimates
about foreign governments, not its own. Tenet, however, did agree reluctantly to do a rushed National
Intelligence Estimate on Iraq’s WMD capability. This intelligence work was undertaken in the wake of Bush’s
and Cheney’s high-profile conclusions on the subject—the vice president’s August 26 declaration that, “Simply
stated, there is no doubt that Saddam Hussein now has weapons of mass destruction,” and the president’s
remark a month later that, “The Iraqi regime possesses biological and chemical weapons.”
The National Intelligence Council, a group of representatives from the key intelligence agencies, began
sifting, sorting and assessing the raw intelligence. The council includes the CIA, the National Security Agency,
which does communications intercepts, the Pentagon’s Defense Intelligence Agency, the State Department’s
intelligence bureau, the Energy Department’s intelligence arm, and the National Imagery and Mapping Agency,
which performs satellite and other overhead reconnaissance.
The group had a massive amount of material, much of it old and not very reliable. Iraq was still one of the
hardest intelligence targets. Saddam had improved his methods of deception and hiding his weapons
rograms—whatever they might be—underground. CIA human intelligence inside Iraq was still weak, and
aramilitary teams like those headed by Tim in northern Iraq had found nothing.
A National Intelligence Estimate is just that, an estimate. During the Cold War it became the document of
choice because it was designed to give the president and his national security team an overall assessment of the
capability and intentions of real threats, such as the Soviet Union and China. NIEs often include political
assessments of the endurance of, say, Colonel Qaddafi in Libya, the direction of the Balkans, famine in Africa,
the chances of war on the Korean peninsula or a nuclear exchange between India and Pakistan.
The format is designed for busy policy makers. So a long NIE of 50 or 100 pages has a kind of executive
summary at the front called “Key Judgments” in which the intelligence analysts would try to give a bottom-line
answer. Would Castro be overthrown? Would Syria attack Israel? Would the Communists win in Nicaragua?
Over the decades there had been much criticism of NIEs by policy makers—and presidents—because the
authors hedge and because the “on-this-hand, on-the-other-hand” reports are littered with maddening
qualifications. No matter what happened, someone could find a sentence or phrase in the NIE that had covered
such a possibility.
Stu Cohen, an intelligence professional for 30 years, was acting chairman of the National Intelligence
Council when the Iraq assessment of WMD was being prepared. He confided to a colleague that he wanted to
avoid equivocation if possible. If the Key Judgments used words such as “maybe” or “probably” or “likely,” the
IE would be “pablum,” he said. Ironclad evidence in the intelligence business is scarce and analysts need to be
able to make judgments beyond the ironclad, Cohen felt. The evidence was substantial but nonetheless
circumstantial; no one had proof of a vial of biological agents or weapons, or a smoking vat of chemical warfare
agents. Yet coupled with the incontrovertible proof that Saddam had had WMD in the past—U.N. weapons
inspectors in the 1990s had found it, tested it and destroyed it—the conclusion seemed obvious.
The alternative view was that Saddam didn’t have WMD. No one wanted to say that because so much
intelligence would have to be discounted. The real and best answer was that he
probably
had WMD, but that
there was no proof and the case was circumstantial. Given the leeway to make a “judgment,” which in the
dictionary definition is merely an “opinion,” the council was heading toward a strong declaration. No pablum.
Analysts at the CIA had long discussed the issue of avoiding equivocation. At times, many, including
John McLaughlin, felt that they had to dare to be wrong to be clearer in their judgments. That summer
McLaughlin had told the NSC principals that the CIA thought they had a pretty good case that Saddam had
WMD, but that others would demand more direct
roof. The CIA did not have an anthrax sam
le, and didn’t