In the dispute over NSA spying, a question that goes to the heart of the matter.
So far, the question “Is is worth it” has been framed by absolutes: on one side, civil libertarians argue you trade freedom for security and get neither, and the pragmatic crowd counters with claims that NSA surveillance has ‘helped’ stop more than 50 terrorist attacks.
The emerging question, really, one that’s been lurking in the numbers all along, is: given the size of the NSA’s program, how do you deal with the problem of false positives? A quick Google search shows a number of cuts at the topic, including this one, which concludes – reasonably enough to my eye – that a positive finding is only likely to yield a real terrorist once in every 10,000 times.
Blogger Corey Chivers notes
While former NSA analyst turned whistle blower William Binney thinks this is a plausible estimate, the point here is not that this is the ‘correct probability‘ involved (remember that we based our calculations on very rough assumptions). The take away message is simply that whenever the rate of an event of interest is extremely low, even a very accurate test will fail very often.
The argument from the pro-NSA spying side is of course we do further sorting once we get a positive result from the spy program. That’s the point of the program, to give us a starting point to work from. What seems impossible is the sheer magnitude of the task – if you have that many potential terrorists, how do you sort the signal from the noise? Looked at that way, calls to limit the government’s spying authority to cases in which there is reason to suspect terrorism may be both the right thing to do, and the practical thing. In any event, there should be some kind of clear explanation of how the government is getting around the false positive problem, if it is.
(The Wall Street Journal has an article with some thoughts on just that, worth reading.)