Has anyone in government done a cost-benefit analysis on bulk surveillance? I mean to say, we’re spending fortunes on this stuff (in the US something like $100B a year ). Does anyone have any idea of whether it’s really worth it? Could we be spending all that dosh more wisely and getting better anti-terrorist results?

Which is why I found this exchange between a questioner and William Binney, the former Technical Director of the NSA fascinating.

Question: Other than making money off of like these NSA contracts, what capabilities do these companies [i.e. defence contractors like Booz Allen Hamilton — Snowden’s employers] *have, what other value are they generating for themselves?

William Binney: Nobody does return on investment at NSA. They don’t.

I mean, if they did return on investment, they would throw away everything except TRAFFICTHIEF and maybe some graphing programs out of MAINWAY– they’d throw all of this away. They wouldn’t have built Bluffdale [Utah], that $2.3B or whatever it is– facility to store data. This is all the data from PINWALE and MARINA and all that stuff is going out there, being stored. So they wouldn’t have to buy that at all. They’d be more effective, because they wouldn’t be buried. So at any rate, that’s what they’re doing.

The exchange comes from an absolutely riveting report of a presentation that Binney gave in which he explained some of the Snowden material.