Wednesday, March 03, 2004

Weapons of Mass Destruction

Roger Simon and reader SR link to an article by Douglas Hanson in the American Thinker. Hanson is not convinced that the methodology employed by David Kay to find Saddam's WMD stockpiles was sound. His doubts are anchored on the shortcomings of Kay's sampling technique:

In his recent testimony, Dr. Kay pronounced that there are no large stockpiles of WMD. This is a pretty bold assertion considering that actual surveys of sites we were familiar with were haphazard and uncoordinated. Also, according to his own interim report published in October of 2003, the ISG had not even searched 120 of the 130 known ammo storage points, much less any underground sites. In addition to these known sites, “neighborhood” arms caches are discovered all the time in Iraq. It is entirely possible that WMD stockpiles were moved out of Iraq, or that they were dispersed in Baghdad neighborhoods and throughout Iraq. All of this may even have been accomplished while the unfocused search operations were ongoing.

Hanson correctly understands that WMDs cannot be regarded as scattered across storage areas according to a normal or other type of distribution. Their locations would tend to be highly concentrated in a few facilities, perhaps even one site. A random sample of a highly stratified phenomenon is unreliable when not all the strata are sampled. To use an analogy, police searching for hidden firearms in a house cannot reach a reliable conclusion after opening 10 of 130 known cabinets on the premises. They would have to search all the known cabinets and look for the unknown ones too. Hanson then goes on to disparage Kay's reliance on debriefing Iraqi scientists.

When Dr. Kay arrived, he shifted the focus from the list of sites to interrogating scientists; not just certain scientists based upon a focused plan, but any and all scientists, as the developing trail would lead. It was apparent that the ISG was largely conducting a massive collection exercise without an operational search scheme to guide it.

This is somewhat unfair. Given the difficulty of searching every site Dr. Kay may have felt that interviewing Iraqi scientists gave him a better chance to uncover the truth. Unlike the location of actual WMDs, the knowledge of their development programs would be presumed (more on this later) to be more widely distributed. A WMD program of any size would presumably involve thousands of technical persons, some of whom may have been unaware of their roles. Kay must have calculated that a careful debriefing of scientists would have made it impossible to conceal the existence of a conspiracy to produce WMDs, however carefully hidden because an account of their activities would inevitably indicate whether they were working on something extraordinary. Yet even that would not be foolproof. An excellent account of Britain's Special Operations Executive research and development activities during the Second World War shows that many SOE personnel remained unaware who they were actually working both during and for many years after the conflict. The SOE's use of common industrial parts often made it unlikely to deduce even what they were finally making. Moreover, the interview methodology suffers from the flaw that unless corroborated independently the interviewer will hear precisely what he wants to hear. Hanson is on much surer ground when he says:

Fear of reprisal from Baathist Party “dead-enders” and enforcers was another very powerful inducement to lying and covering up important information. Lacking corroborating documents to trap liars, scientist interrogation became another collection effort with no strategy for identifying and checking on the veracity of key personnel.

Yet the major difficulty remains overlooked. The key problem to finding Iraqi WMDs comes from the underlying assumptions about how large a national WMD program had to be. In the classic Cold War model, any WMD program was assumed to be very large, a copy of the Manhattan Project. Following from that assumption, the tools used to detect these programs were overhead imaging, environmental sampling and debriefing technical personnel. These tools did an excellent job at finding centrifuges, reactors, weapons storage facilities and their associated delivery platforms. But the rollup of the A. Q. Khan network, the "Johnny Appleseed" of nuclear proliferation, in the New Yorker's felicitous phrase, showed the world another model, and in the context of Iraq, the more likely model of WMD development. In this model, the only relevant WMD manufacturing facility is a pile of cash. Everything, including possibly the fissile material, was potentially for sale from A. Q. Khan's worldwide network. From the Seymour Hersh's New Yorker article:

Robert Gallucci, a former United Nations weapons inspector who is now dean of the Georgetown University School of Foreign Service, calls A. Q. Khan "the Johnny Appleseed" of the nuclear-arms race. Gallucci, who is a consultant to the C.I.A. on proliferation issues, told me, "Bad as it is with Iran, North Korea, and Libya having nuclear-weapons material, the worst part is that they could transfer it to a non-state group. That's the biggest concern, and the scariest thing about all this-that Pakistan could work with the worst terrorist groups on earth to build nuclear weapons. There's nothing more important than stopping terrorist groups from getting nuclear weapons. The most dangerous country for the United States now is Pakistan, and second is Iran." Gallucci went on, "We haven't been this vulnerable since the British burned Washington in 1814."

Therefore the lower size bound of an Iraqi WMD program was no facilities at all. An Iraqi WMD program only had to be as big as Al Qaeda's. Saddam may have simply decided it was cheaper to buy a weapon or near-final components instead of building them from scratch. The assumptions about the character of Iraq's possible WMD programs bear directly on the failure of prewar Allied intelligence to characterize and describe the program accurately. If the relevant model was not a cheap version of the Manhattan Project but rather the A. Q. Khan commodity model, it would have misled the analysts seriously and caused them to overlook the one alarming corollary. Every WMD component in the A. Q. Khan manufacturing chain had a tradeable value. In that universe, nuclear, chemical and biological weapons are not merely instruments of state power but fungible financial assets. Saddam would have looked at a nuke or bioweapon not simply as a lethal device but as an investment. Dr. David Kay's findings may not mean that Saddam destroyed or hid his weapons before the war. It may merely mean that he sold them.


Reader J2 links to the Foresight Exchange, where bets are traded on a variety of subjects including hypothetical terrorist events.