CCTV fails to cut crime - so let's take identity out of the equation

A senior police officer has stated that the pervasive use of CCTV in the UK has failed to cut crime and is an ‘utter fiasco’. with only 3% of London’s street robberies being solved using security cameras.

I considered some alternative headlines for this article – ‘Pope reaffirms commitment to Catholic faith’ and ‘bears decide not to install indoor toilet facilities’ were possible candidates.

At a recent conference, a former senior police officer (Chatham House Rule so I won’t identify him) reflected on his force’s introduction of CCTV in a busy city centre. As one of the first such systems in the UK, it was very effective at tackling antisocial behaviour, by allowing the control room to send officers immediately to the scene of the disturbance.

In such applications CCTV is extremely useful. It’s there to identify the crime – not the criminal – and ensure that the perpetrator is caught in the act. But the tragic Jamie Bulger case changed public attitudes. Dramatic images of two young boys leading a toddler to his death catalysed the rollout of CCTV around the nation with little thought of how imagery might be used.

Last year I waded in to stop a mugging on a late evening train home (which wasn’t as heroic as it sounds, the elderly gentleman being attacked for his iPod was demonstrating some commendable boxing skills, and if his assailants hadn’t been so stoned that they felt no pain then neither would have got up after the punches he landed). The train guard informed us that the incident had been caught on CCTV and the control room could have the images with the police within minutes. The victim quite predictably declined to pursue the case, since the attackers had long since fled, and what was the likelihood that an overstretched police force was actually going to do anything other than waste everyone’s time with a stack of paperwork?

Our problem with CCTV now is the fact that there is so much of it about, and we’ve forgotten why we wanted it in the first place. First we used it to identify a crime, then the criminals, now we’re trying to spot criminal behaviours before crimes are even committed. Various trials of facial recognition and gait recognition schemes seek to spot known offenders when they enter the field of view, and to track their movements from camera to camera. But these technologies are no substitute for trained eyes on monitors, and there simply aren’t sufficient operators available for the number of cameras we have. Police are inundated with CCTV data and it is becoming worthless in all but the most major criminal investigations. Even the London tube bombers were only found on CCTV many days after the event.

It’s high time that we move the focus of our debate from ‘how much CCTV is acceptable?’ to ‘what are we going to use it for?’ When we stop trying to use it to identify criminals (and by implication criminalise anyone the camera sees), and instead apply it to spotting crimes in progress, we’ll have a police tool that might actually be useful. Without the pressure to identify subjects, operators can focus on using CCV for a constructive purpose, and the rest of us can stop worrying about one key aspect of its invasion of our privacy.