The Effectiveness of Public CCTV
John has synthesized a number of the papers on CCTV’s effectiveness on crime. His work is careful and measured on a topic that is highly polarizing. Its a must read and I share many of John’s views. He has also been very helpful in providing links to the papers he reviewed. At some point, I would like to do a detailed post on his review (as well as see if I can add more papers). But for now, here are some of the key findings.
The expectation that CCTV systems should be deployed to reduce crime rather than solve crime has created huge problems. –
While the studies show serious doubt on CCTV’s ability to reduce crime generally, a strong consensus exists in CCTV’s ability to reduce premeditative/property crime
CCTV is consistently treated as a singular, stable technology, obscuring radical technological changes that have occurred in the last 10 years
Routine comparison of police vs cameras is counterproductive
This leads him to conclude:
Stop claiming that CCTV can generally reduce crime
Optimize future public CCTV projects around crime solving rather than crime reduction
Optimize future public CCTV projects around material and premeditative crimes
Hi Rajiv,
Thanks for the post and feedback.
One aspect I did not cover was video analytics. Given the focus of your blog, I wanted to briefly share my thoughts on this aspect.
Conceptually, video analytics offer the ability to increase public CCTV effectiveness in reducing crime. The video analytics should give real time alerts on incidents in progress which should lead to police intervention and a deterrent effect.
I did not recommend video analytics because I am concerned about how well it will work. Part of the issue is the effectiveness of the analytics but another significant element is how quickly a responder could arrive on scene. Unlike private systems where responders are often located on site, with public scenarios it could take 5, 10 minutes or more for a responder to be dispatch. My third concern is that video analytics tend to work best on clearly differentiable phenomenon like jumping a fence. A lot of public systems need to handle altercations, drug deals, etc – things that video analytics usually has challenges in performing.
I do think video analytics might enhance existing active monitoring schemes but I think more strongly that active monitoring schemes should be reduced.
In any event, as video analytics mature, I do think they can be valuable for public surveillance but I am concerned about them creating another black eye for security.
Best,
John