A UC Berkeley report on whether security cameras have deterred crime in San Francisco is
conclusively inconclusive,
according to the mayor.
Schneier also commented on the article; unsurprising, as the security camera is a popular target of
his security theater argument.
I just finished reading
Robert A. Heinlein's Friday, where he speculated that, in the future, monitoring security cameras in 'real-time' would require sampling, in practice; he implied that this was common knowledge, at least amongst criminals, and that they managed their risk accordingly (e.g., they went ahead and played the odds, or destroyed the camera, etc.).
However, the reality that this and other articles reinforce is that all of these decisions - including whether to install security cameras, how many to install, how to operate them, and how to monitor them - are heavily influenced by money. If there's money available, why not install security cameras, just in case, in other words; this, despite the fact that
compelling arguments against the collection of unnecessary data have been put forth since the 60s.