CCTV Prevalence in Britain

For many years the British public has often been told that the United Kingdom has 4.2 million CCTV cameras—that’s one for every fourteen residents—as widely quoted by politicians, various media, and even the police.

This statistic is rarely questioned, but thanks to a recent episode of the excellent More or Less (UK-only?) suggesting that this statistic was, at best, dubious, I decided to do some reading.

I didn’t have to read much.

The statistic comes from a 2002 report from the URBANEYE project, looking at the prevalence of video surveillance in London (pdf). From the Conclusion:

In our Putney sample, 41% of premises had CCTV systems in operation. These institutions had an average of 4.1 surveillance cameras. If we use these figures to extrapolate the extent of CCTV coverage in London and the country as a whole we come up with the following results. If we begin by assuming that the extent of CCTV coverage in Putney is broadly representative of CCTV coverage across the whole of London, we could estimate that 41% (102,910) of the 251,000 business registered for VAT in London would have a CCTV system. Between them these businesses will have 421,931 surveillance cameras. If we add to these the number of surveillance cameras operating in other public institutions (open-street systems, transport, hospital, schools etc.) it would not be unreasonable to ‘guesstimate’ that Londoners are monitored by at least 500,000 CCTV cameras. This means that in London (with a population of 7.2 million residents) there is approximately one camera for every fourteen people. From these figures we would suggest that in the UK (with a population of almost 60 million) there are at least 4,285,000 cameras in the UK.

The Putney sample was a paltry 211 premises. And Putney, as one of the 35 major areas in Greater London, is hardly representative of the UK as a whole. Even the CCTV User Group says the results are “extremely questionable”.

3 thoughts on “CCTV Prevalence in Britain

  1. range

    Sample is way too small to be relevant. You’d need multiple samples in different cities, with a lot more samples. This is basically insignificant.

Comments are closed.