The interesting John De Palma writes to me:

In response to the comment you made beneath your Iraq War blog entry&#8230-

Wolfers/Zitzewitz wrote: &#8220-&#8230- the public information on the probability of weapons of mass destruction in Iraq appears to have been of dubious quality, so it is perhaps unsurprising that both the markets were as susceptible as general public opinion to being misled.&#8221-
(The paper is posted here &#8211-PDF file. Figure 5 charts the historical pricing of the TradeSports WMD contracts.)

Also, Wolfers/Zitzewitz/Snowberg wrote in a separate paper: &#8220-Figure 5 shows the price of a contract on whether or not weapons of mass destruction (WMD) will be found in Iraq. Note that at some points the value of the contract exceeded 80%, yet weapons were never found. It is likely that this market performed poorly since the cost of gaining new information was quite high. Since WMD can be non-existent almost everywhere, but still exist somewhere, it was difficult to bet against the strong case made by the White House, at least initially.&#8221- Paper: PDF file.

An article in the NY Times last year reminded me of the TradeSports WMD market. From the New York Times:
&#8220-The Iraqi dictator was so secretive and kept information so compartmentalized that his top military leaders were stunned when he told them three months before the war that he had no weapons of mass destruction, and they were demoralized because they had counted on hidden stocks of poison gas or germ weapons for the nation&#8217-s defense.&#8221-
This assertion in the New York Times article suggests that even Iraqi &#8220-top military leaders&#8221- would have been on the wrong side of the TradeSports WMD trade.

A few years ago when Prof. Wolfers spoke at Columbia Univ, he used the WMD market as an example of an inefficiently priced one, consistent with what he wrote in those papers. It seemed to me that the footprint of what he was saying could be in contracts of that nature being priced too high. (Though I would think it difficult to empirically support that point since even if contracts like that are priced too rich historically, it could reflect the necessary posted margin involved with going short.)

Thanks.

From the Wolfers/Zitzewitz paper: Prediction Markets – 2004 – PDF file