Just because you're paranoid doesn't mean they aren't out to get you.

Floyd Rudmin, a professor in Norway, in an online essay titled The Politics of Paranoia and Intimidation, argues that it is statistically impossible for the NSA mass-surveillance program (of American e-mail and phone calls) to detect terrorists effectively, since the false-positives of such an effort would vastly outnumber the actual terrorists you could find, unless there are a million or so "terrorists" as the target of your search.

It's more likely that the program has political motivations.

The ratio of real terrorists to innocent people in the prison camps of Guantanamo, Abu Ghraib, and Kandahar shows that the US is paranoid and is not bothered by mistaken identifications of innocent people. The ratio of real terrorists to innocent people on Bush's no-fly lists shows that the Bush administration is not bothered by mistaken identifications of innocent Americans.

Also, mass surveillance of the entire population is logically plausible if NSA's domestic spying is not looking for terrorists, but looking for something else, something that is not so rare as terrorists. For example, the May 19 Fox News opinion poll of 900 registered voters found that 30% dislike the Bush administration so much they want him impeached. If NSA were monitoring email and phone calls to identify pro-impeachment people, and if the accuracy rate were .90 and the error rate were .01, then the probability that people are pro-impeachment given that NSA surveillance system identified them as such, would be p=.98, which is coming close to certainty (p=1.00). Mass surveillance by NSA of all Americans' phone calls and emails would be very effective for domestic political intelligence.

Here is the article at: http://www.lewrockwell.com/orig7/rudmin1.html
Found via a comment posted at Conscious Junkyard. The comment poster also refers to a Bloomberg article reporting a claim in court that the surveillance program pre-dates its supposed cause: the 9/11 terror attack.