More Wiretap Statistics

The secrecy (necessary) of the FISA and warrant process makes it very difficult for an average citizen to assess the Bush Administration's claims. But one thing that helps at least assess the claim that the sheer volume is overwhelming is some sense of the scale involved. In a recent post I cited a link that gave at least some information:

(1) The very gold standard of a terrorist call was one from Bin Laden's own satellite phone, until he stopped using it in 1998. During the preceeding two years, he made 1,100 calls. The recipients of 100 of those calls remain unknown.

(2) Of the 1,000 calls traced, over 200 were to Britain, but only 27 phone numbers were involved.

(3) Six (6) of those calls were to the United States.

(4) Bin Laden's phone was only the tip of the iceberg. There were numerous other Al-Qaeda communications locations, especially a hub in Yemen (since shut down).

The total phone numbers captured from Al-Qaeda documents, laptops and cell phone in the Afghan War presumably dwarfs the number called by Bin Laden. But it need not be so overwhelming, since a single number can receive multiple calls. Projecting from Bin Laden's calls, I postulated that if six Al-Qaeda contacts out of 1,000 are with the US, it would take 10,000 phone numbers to amount to 60 in the US and 100,000 numbers to reach 600 US numbers. How many warrant applications can the FISA Court handle?

An article from the Los Angeles Times (H/T Talking Points Memo) offers some statistics about FISA warrants versus prosecutions since 1998:

1998: 796 warrants; 69 prosecutions (ratio 11.5 to 1)
1999: 880 warrants; 99 prosecutions (ratio 8.8 to 1)
2000: 1,012 warrants; 75 prosecutions (ratio 13.5 to 1)
2001: 934 warrants; 115 prosecutions (ratio 8 to 1)
2002: 1,228 warrants; 1,208 prosecutions (ratio nearly 1 to 1)
2003: 1,724 warrants; 899 prosecutions (ratio 2 to 1)
2004: 1,754 warrants; 762 prosecutions (ratio 2.3 to 1)
2005: 2,072 warrants; 771 prosecutions (ratio 2.7 to 1)
2006: 2,176 warrants; 624 prosecutions (ratio 3.5 to 1)
2007: 2,370 warrants; 505 prosecutions (ratio 4.7 to 1)

We do not know the origin of all these new warrants, but the number nearly tripled since 1998. Presumably at least some are the result of information captioned from Al-Qaeda in during the war. Indeed, the total increase is considerably larger that the 600 numbers I have speculated as the maximum number of US numbers we could have expected to capture overseas, but FISA appears to have handled them.

At least as spectacular as the rise in number of warrants is the relative increase in number of prosecutions. The article focuses on the rising number of warrants, even as prosecutions have been falling. The author wonders if this means that many of these wiretaps lack merit. Any answer would be speculative, but let us consider. The totals are not limited to Al-Qaeda, but include all national security warrants and prosecutions. At least some presumably are long-term wiretaps of foreign embassies, consulates and diplomats that can never result in prosecution because of diplomatic immunity. And some no doubt involve spies for foreign governments or non-Islamic terrorists, again, probably not much changed since before 9/11. But it seems safe to assume that the increase in warrants and prosecutions mostly involve Islamic terrorism.

Government officials quoted in the article defend wiretaps that do not result in prosecutions. They point out that there are always at least some false leads, and that sometimes when evidence is not sufficient for prosecution, terrorist plots can be thwarted by other means, such as "military or diplomatic pressure," deportation or simply letting potential conspirators know they are being watched. And, although the article does not mention it, terrorist plots, after all, are usually conspiracies with multiple members. Investigating a conspiracy may involve numerous wiretaps but yield only one prosecution of multiple defendants. In short, there is not sufficient evidence, and the FISA Court is not at liberty to release sufficient evidence, to evaluate how many of the wiretaps do or do not have merit.

But the article misses a much more obvious point. To this day, there are far more prosecutions relative to wiretaps than before 9/11. In all of the years listed before the September 11 attacks (including 2001, the year of the attacks), there were at least eight wiretaps to one prosecution. That number fell to almost as many prosecutions as wiretaps in 2002. The ratio has been moving closer to pre-9/11 norms since, but still has not reached them.

Unlike FISA warrants, national security prosecutions are matters of public record. How many of them have been for Islamic terrorism and how many for other offenses? How many prosecutions were of a single defendant and how many for multiple. How immediate was the threat? (The article lists an example of a sailor convicted of passing classified information to a pro-terrorist website. That sounds legitimate, but not an urgent danger. Prosecutions of alleged conspiracies to commit actual terrorist acts have been notoriously dubious. Any many terrorism prosecutions are for fundraising). And what have the dispositions been of these cases? Without this further information, it is not possible to answer more obvious question raised by the article's statistics -- is the government regularly prosecuting terrorism cases that lack merit?