CCTV: national standards to make the surveillance work

CCTV: national standards to make the surveillance work

Contents at a glance

Research from the Home Office suggests that many CCTV systems have been procured without detailed consideration as to the required functionality that makes them fit for purpose. According to official documents, 80% of CCTV images cannot reliably be used by the police.

The guidance issued by the Home Office to assist those public authorities purchasing CCTV systems has been largely ignored. This guidance states the CCTV procurement process should involve; a detailed specification of processing objectives for CCTV, a multi-disciplinary approach to identify the social problems which CCTV will help to reduce, a comprehensive analysis of technical and managerial matters which identify how CCTV will be used in practice, and a business case which identifies why CCTV implementation is the best solution.

CCTV technology has changed from one based on sequentially-recorded tapes to a technology based on digitised images processed by sophisticated software; images of individuals caught on camera unambiguously fall within the remit of the Data Protection Act.

The Information Commissioner's draft Code of Practice on CCTV stresses that the technical choices made at the procurement stage determine certain privacy and surveillance outcomes; a privacy impact assessment is needed to justify the procurement of a CCTV system.

The Home Office and the Association of Chief Police Officers have published a National CCTV Strategy that will, in future, determine standards and procedures that have to be adopted by those operating or procuring digital CCTV systems. In urban areas, further expansion and integration between existing CCTV systems is the likely outcome if the Strategy is implemented.

The Strategy suggests that legislation might be needed to create a separate body to supervise and enforce digital CCTV standards; it is possible that this body could become responsible for the supervision of privacy standards with respect to CCTV operations.

Introduction

In October, the Home Office and the Association of Chief Police Officers (ACPO) published “A National CCTV Strategy” (the “Strategy”) designed to apply common standards, procedures and technical platforms to all public sector CCTV systems and to many private sector CCTV systems. The publication of this Strategy coincided with the closure of the consultation period associated with the Information Commissioner's consultation on a revised “CCTV data protection code of practice”. This code is to replace the existing code (published in 2000) and the CCTV guidance issued in the wake of the Durant judgment (published in 2004). As well as raising the expected data protection issues, the Commissioner's draft Code of Practice (the “draft Code”) also dwells on technical and procedural matters.

These technical matters have become important because CCTV systems are now based on digital images processed by computers and not the sequential recorded images that one usually associates with VHS tapes. As we shall see, the choice of that digital technology can raise specific data protection issues and a purpose of this article to explain why this is the case. In our view, the operation of your organisation's CCTV system is something that you will need to consider in the near future.

“Looking out for you” - some CCTV history

In 1995, the then Home Secretary, Michael Howard MP, published a document “CCTV - Looking out for you”. It emphasised that CCTV should not become “a cure looking for an illness” and stated that “too often CCTV is perceived as the 'cure' before clearly identifying the problems which it is supposed to deal with”. To illustrate these issues, the document contained two scenarios – both involving local authority proposals to install a CCTV system.

The first scenario related to one (anonymous) borough where there appeared to be a high rate of vandalism to cars, and this level of crime “was believed to be driving people from using the town centre” and it was thought that “CCTV would be the best solution”. However, research showed that “the damage was concentrated on particular days of the week” and “most commonly appeared during one period of time on those days". Further investigation showed that the higher rate of damage was “caused by market traders when they set up early in the morning” and that implementation of a “town centre CCTV would have been an expensive non-solution”.

The second case study concerned the CCTV system installed at Newcastle-upon-Tyne City Centre. Here a police superintendent explained that, prior to the installation of CCTV, a “revised style of pro-active policing” was introduced. This resulted in a decrease of crime from 13,500 incidents in 1991 to 4,500 in 1993 and the introduction of CCTV reduced crime statistics further to 2,381 offences per year. The conclusion drawn was that “the true success of CCTV” was linked “to a strategy of pro-active policing and other positive initiatives which are inextricably linked to the use of CCTV”.

The Home Office guidance then outlined the basic components of a successful installation of a CCTV system; ingredients that are repeated in the Commissioner's draft Code and ACPO's Strategy. These components, which provide a useful checklist of actions to be taken prior to purchase, involve:

·        a multi-agency approach to produce a CCTV system specification that involves all stakeholders likely to have an interest in the effectiveness of the system (e.g. police, local authorities, prosecutors, car park operators, residents groups, media representatives and representatives of retailers, restauranteurs and publicans). This approach will lead to the identification of the common social problems that CCTV aims to reduce (e.g. vandalism, shoplifting, drunkenness, burglary, pick pocketing, disorderly behaviour) and to an assessment of alternative solutions that don't need a CCTV system installed (e.g. better lighting, neighbourhood watch, improved communications with police, security patrols, no tolerance to graffiti, bye-laws re drinking, barrier entry car parks).

·        a clear understanding of how and why the procurement of a CCTV system would be the best solution to mitigate the impact of the identified social problems (e.g. by providing a deterrence to the targeted behaviour or evidence that can be used in Court), and how CCTV will be co-ordinated with other initiatives (e.g. policing and anti-social behaviour strategies).

·        the identification of the technical requirements from the system specification. This ensures that a procured CCTV system would have the functionality to deliver measurable results in relation to the identified problems (e.g. the system operates under the required lighting conditions; cameras have the correct zoom functionality; colours are reproduced accurately; evidential issues are resolved by the correct choice of monitors, cameras or storage devices).

·        a business case that specifies the outcomes or deliverables to demonstrate that the CCTV system will be (or continues to be) a cost-effective solution to the identified social problems.

·        careful consideration of staffing and operational issues (e.g. management, control room procedures, staff training, disclosure procedures, subject access procedures) and public relations issues (e.g. reports to Councillors, fair processing notices; press and public liaison) to maintain public confidence in the system.

“Assessing the impact of CCTV”

The 1995 guidance from the Home Office appears to have been ignored in the rush to spend the £210 million the Government made available for the capital spend on CCTV systems between 1994 and 2003. This conclusion is apparent from a Home Office Research Study (“Assessing the impact of CCTV”, Report no. 292, Feb 2005) which studied 13 CCTV systems that were installed during this period.

Although the Home Office research contained a caveat not to come to “too simplistic a conclusion” about the effectiveness of CCTV in general, there is no mistaking its central message. The research report notes that: “It would be easy to conclude from the information presented in this report that CCTV is not effective: the majority of the schemes evolved did not reduce crime and even where there was a reduction, this was mostly not due to CCTV; nor did CCTV schemes make people feel safer, much less change their behaviour”.

This rather frank statement is accompanied with pithy explanations in relation to the management and procurement of CCTV systems studied by the researchers.

• “Many projects did not have clear objectives” and that the general perception was that “CCTV was a good thing”. Sometimes installation of a CCTV created “demands by neighboring towns to catch up” and the “existence of funding for CCTV created pressure to bid for it”.
• “Many schemes relied too heavily on technical consultants whose work was not scrutinised, largely because no one had the qualification to question what was done”. Since “a consultant was dispensed with in many cases, planners were unable to challenge the technical sales pitch of equipment suppliers”.
• “Some systems failed to engage properly with end-users, most notably the police” and this resulted in “a loss of interest in the system and a reluctance to use the evidence supplied by the cameras”.

The research also noted that “there was a lack of realism about what could be expected from CCTV” and that “in short, it was oversold – by successive governments – as the answer (indeed magic bullet)”. The researchers commented that “few seeking a share of the available funding saw it necessary to demonstrate CCTV's effectiveness” and “it was rarely obvious why CCTV was the best response to crime in particular circumstances”. There was “a tendency to put up cameras and expect impressive results, ignoring the challenge of making what is quite a complex measure work, and failing to define what exactly the CCTV system was expected to do”.

The research also noted positive items about CCTV. For example, in the 13 systems evaluated, crime had decreased in 6 areas whereas the increase in 7 areas covered by CCTV could not be attributed to the CCTV system. CCTV had less effect on impulsive crime (e.g. alcohol-related) but had a measurable effect on premeditated crime (e.g. car theft) and there was evidence to suggest that CCTV could assist where specific problems had been identified (e.g. areas where drug dealing was prevalent or which were associated with “acquisitive crime” such as theft). Fixed areas with specific problems (e.g. car-parks) were generally a CCTV success story.

The research noted that technical issues such as camera coverage, lighting and location were important to the effectiveness of the CCTV system but not if camera density had reached saturation point. Members of the public worried less about crime in the areas covered by CCTV, but “knowing that cameras were installed in the area did not necessarily lead to reinforced feelings of security among respondents”. Above all, procurement of a CCTV system which had been integrated with other crime reduction policies (and which involved police, retailers, community wardens, publicans etc) worked when installation included consideration of the components identified in the 1995 guidance (and listed above).

The rush to digital technology

At our Update sessions (October 2007), we have been fortunate to have a guest speaker, Bernie Brooks, Founder of Datpro Ltd, a company that specialises in CCTV consultancy. He presented a liturgy of technical issues, many of which echo the Home Office research findings and the 1995 guidance, to arrive at a simple message. The intended use of a digital CCTV system and its operational framework has consequences for the choices made when the technology is purchased; and a poor technical choice can make a CCTV system unfit for its stated purpose.

Some of the main issues identified by him included the following.

• End users do not know what they want from their CCTV; they often follow supplier recommendations and do not have the technical expertise to define what they want.
• Procedures and standards in relation to security (e.g. Code of Practice on Information Security management ISO27001/ISO27002) or evidence (e.g. BS DISC PD0008: 2004 Legal Admissibility and Evidential Weight of Information Stored Electronically) need to be integrated into the operational procedures associated with a CCTV control room.
• Issues such as camera functionality, camera location, image quality, image resolution and camera density interact with technical aspects of recording (e.g. disk capacity, frames per second) that are not usually considered by purchasers of CCTV systems. For example, a standard 350GB storage unit (often associated with CCTV technology) can store the images from a single camera, running for 21 days, recording at 5 frames per second. As most of our readers watch terrestrial TV (which operates at 25 frames per second), it can be seen that a 350GB disk would only hold 4 days of TV quality images.
• If the recording hardware handles more than one camera, then the maximum recording rate might have to be shared between all cameras on a CCTV system. Suppose there is a CCTV system that can record at a 25 frames per second rate and the computer supports 10 cameras. If one camera is recording at 18 frames per second, it follows that the other 9 cameras will be recording at 1 frame per second. This rate of recording impacts on the evidential value of the stored images.
• Retention periods associated with CCTV images have, possibly as a result of poor choice of recording technology, fallen and a 14-day retention period is now commonplace (rather than 28 or 31 days, preferred by the police). Movement activated cameras can be used to overcome recording problems (e.g. cameras only record events when something happens), however, such movement cameras can be triggered by snow, or trees swaying in the wind.
• Operational problems have arisen because there has been a failure to consider how images are to be shared with the police or prosecutors. Often images cannot be used as evidence. Sometimes the images the police need cannot be isolated and this runs the risk that the police might have to seize the computer.

Commissioner's Code of Practice consultation

Given the above, it can now be seen why the Commissioner in his draft Code is asking questions such as: “for whose purposes will the CCTV be used?”; “What are the problems it is meant to address”; and, most importantly, “can CCTV technology realistically deliver”? This is because several data protection issues arise directly from the incorrect technical functionality and the implementation of an ill-considered CCTV system specification.

For example, internet systems of control for a CCTV system could be vulnerable to hackers who could then enable or disable functionality of the system (Seventh Principle). Cameras are so small that many individuals are unaware of CCTV surveillance (fairness- First Principle); this could give rise to claims of unlawful covert surveillance (unlawfulness- First Principle). If there is a viable alternative to the installation of a CCTV system (e.g. installation of better lighting), or if such systems do not perform to deliver the specified purposes (e.g. images of insufficient quality that can't be used as evidence) then adequacy issues would arise (Third Principle). Digital images are easily manipulated and this gives rise to other security issues (Seventh Principle) and/or data integrity issues (Fourth Principles). Failure to consider the rights of data subjects engages the Sixth Principle.

The Commissioner stresses that data protection also needs to be integrated into day-to-day operational procedures. For example, if there are disturbances in an A&E department of a local NHS trust, should the police be able to take a “live feed” in order to record any relevant evidence and provide a continuous police presence? In this case, the solution proffered by the Commissioner is to install a panic button, which if activated by hospital staff, starts the live feed to the police. In this way, the police do not observe those who visit A&E when there isn't an incident (this would be excessive- Third Principle).

The Commissioner is also concerned to ensure that the data controller is identified so that responsibility for meeting data protection obligations are unambiguously assigned (e.g. so that the rights of data subjects to access to images containing their personal data are maintained). In this latter regard, functionality to blur images of other individuals caught in these images, if needed, becomes an important factor if personal data are to be released following a subject access request. Management of procedures, maintenance of equipment, and staff training (Seventh Principle) are an important element in the system of control, as does the maintenance of evidence that operational procedures are actually followed. Finally, it is estimated by Bernie Brooks that about one million data controllers are not registered for the processing of personal data for the purposes associated with CCTV.

The Commissioner's red-lines

There are four “be careful what you do areas” identified in the draft Code. These are:

• CCTV must not be used to record conversations. The draft Code states that “CCTV must not be used to record conversations between members of the public as this is highly intrusive and unlikely to be justified” and that “You should choose a system without this facility if possible. If your system comes equipped with a sound recording facility then you should turn this off or disable it in some other way”.
• Audio based alert systems (when the cameras are activated by a noise) “may be acceptable subject to sufficient safeguards, but conversations should not be recorded and the operators should not listen in”.
• “Help” and Panic buttons, if activated by those who need help are acceptable, as they are activated by the person requiring assistance. This procedure can be seen as a communication instigated at the consent of the individual seeking the CCTV operator's attention.
• Broadcast messages via speakers fitted to cameras that allow operators to admonish wrongdoers observed on camera in real-time is acceptable (e.g. an operator who says “will the person in the yellow jumper pick up the crisp-packet he has just dropped”). However, “the use of audio to broadcast messages to those under surveillance should be restricted to messages directly related to the purpose for which the system was established” (e.g. an operator who says “I'm confused as to what you are pointing at because you are using two fingers” rather than the broadcast comment “And the same to you too!”).

Strategy to extend the role of CCTV

The starting point for the Strategy is that CCTV is a success story. In this view of the world, the research that has indicated CCTV to be ineffective has merely demonstrated that CCTV specifications that have not been properly defined, do not produce the satisfactory results when the system is implemented. That explains why the Ministerial Foreword to the Strategy concludes that “although the CCTV infrastructure is very good, it could be much better if it was effectively co-ordinated”.

The Strategy sets out plans for co-ordinating an ambitious, integrated expansion of the CCTV in town centres to include “CCTV from buses, tube and train carriages” and from “football stadiums, arenas and other areas of public convenience”. The Strategy foresees other electronic linkages for localised CCTV systems such as in a store or railway station: these include “shop cameras to Electronic Point of Sale systems”, “transport system cameras to travel cards” and “internal building cameras connected to building access control systems”. Such integration, the Strategy states, will “dramatically improve the effectiveness of CCTV systems” as “post event CCTV images can quickly be searched against other events”.

One important reason for these linkages is that, in the post 9/11 world, the Strategy is subtly enhancing the role of CCTV from its accepted role that relates to crowds in city centres (e.g. public safety, public order or low-level street-crime) to ensure that such CCTV, in future, has the functionality to trace individuals and vehicles involved in serious crime and for anti-terrorist purposes. The Strategy clearly states that “if we are to deal more effectively with serious, organized crime and terrorism, different operational requirements are needed”.

In the UK news section, we reported that these changes are already being developed. London's Congestion Charge cameras now feed images through to MI5 for national security purposes, and modern digital CCTV in city centres are increasingly augmented by an automatic number plate reader (ANPR) functionality that permits checks with the Police National Computer (e.g. to provide intelligence on vehicle movements, to identify uninsured drivers). The Strategy envisages that CCTV will develop a facial recognition functionality in future, and one can see such systems being used in relation to ASBOs or surveillance of individuals of interest to the police. It can also be anticipated that the digitised facial images collected for the proposed ID Card system, or used in connection with the existing Passport and Driving Licence systems would become a useful source for facial images of most of the population. In this way, the police's use of CCTV, linked to ANPR, to support its policy of “Denying Criminals the Use of the Road” develops into a policy of “Denying Criminals the Use of the Pavement”.

These wider security objectives explain why the Strategy also envisages a series of CCTV networks, where CCTV different systems will become interlinked, and the creation of centralised repositories of images. Although such a repository of images is not technically feasible at the moment, the Strategy envisages building “a skeleton infrastructure and a strategic network of CCTV cameras to assisting in the policing of key economic sites and the investigation of major and terrorist incidents”.

Repeated messages re CCTV failings

The Strategy repeats all the key messages from the 1995 guidance in relation to CCTV procurement (see above) in order to justify the imposition of national, enforceable standards. For example, the Strategy notes that in the days when CCTV footage was recorded to VHS tape, there was compatibility in camera resolutions output specifications, transmission standards and recording quality.

However now digital cameras from one manufacturer may not be compatible with others recording systems and that the Criminal Justice System “struggles to play back images stored in the many different proprietary recording formats”. With the emergence of digital CCTV “the situation has taken a turn for the worse” and over “80% of today's CCTV footage is far from ideal for police purposes”. This means many CCTV systems cannot be used for the detection and prevention of crime (as the public believes) but to “monitor crowds, slips, trips and falls”. The Strategy concludes that “often there is a public expectation that these systems are being installed for their safety, but the CCTV may not be of sufficient quality for police to use in criminal investigations”.

The Strategy stresses that a fundamental issue relates to the purpose of a CCTV scheme. For instance, cameras used in a post-incident investigation have to possess a different functionality from cameras used for pro-active policing. The former needs a fixed camera to ensure continuity of any evidence, the latter needs a roving functionality in order to follow a suspect. The Strategy notes that the dual use of CCTV “frustrates investigators” as details of an incident “are not captured initially” and are “unable to provide secondary evidence/intelligence by providing a good quality continuous recording at set locations or choke points in the vicinity of a crime”.

The regulation of CCTV

The Strategy's plans amount to an expansion of the role of CCTV and the integration of city-centre CCTV systems with other cameras that record other facets of daily life; this increases the importance of regulatory checks and balances. The Strategy recognises the need for improved public accountability but then notes that “whilst there are always likely to be some privacy concerns, these should be addressed within existing laws”.

The Strategy achieves its view of enhanced public accountability by: implementing a licensing scheme for CCTV systems; establishing a CCTV registration scheme that contains basic details of the CCTV scheme; imposing national training programs and standards for CCTV users; extending the application of planning laws to cover CCTV; and creating a stronger role for the Information Commissioner in enforcing CCTV standards.

The Strategy also notes that there is a confused line of regulation at the moment because there are three bodies that could be involved in the supervision of CCTV. These bodies are:

• the Information Commissioner who looks at data protection matters. The Commissioner does not have any remit in relation to licensing CCTV systems or enforcing procedural or technical standards that are not privacy related; in addition, the notification of “registered particulars” under the Act is not designed to cover the registered items identified in the Strategy. If the Information Commissioner's role were to be extended to enforce CCTV standards, then he would become responsible for a technology that is designed to be invasive of privacy; this could compromise his current role which is largely seen as protecting data subjects from invasions of privacy.
• the Surveillance Commissioner whose limited privacy remit largely covers warrants or authorisations in relation to covert and directed surveillance undertaken by the police and the national security agencies. Because this Commissioner has a very specialised regulatory role, it is very unlikely that his role would be extended to include a responsibility in relation to extensive and overt surveillance via CCTV.
• the Security Industry Authority (SIA). The SIA has implemented a licensing scheme for “Public Space Surveillance” but its prime role relates to private security industry (e.g. private investigators and night club bouncers), and not the minutiae of enforcing CCTV standards (e.g. to the public sector). The SIA requires a private body to obtain a licence when CCTV is used to monitor the activities of members of the public in public spaces; however the SIA does not have a role in relation to privacy, although breach of data protection laws, could in theory, result in a loss of a licence to operate a CCTV system.

Is “confusion” to be replaced by a “conflict of interest?”

To remove such confusion between the roles of the various regulators, the Strategy suggests there should be a “body responsible for the governance and use of CCTV in the UK”, and the development of “legislation to ensure the appropriate regulation of CCTV”. Thus, despite the Strategy's reference to widen the Information Commissioner's remit, it is reasonable to expect that any new law specifying a system of governance could also establish a dedicated CCTV Commissioner, rather like the Commissioner created by the ID Card Act. In this way, regulation of the privacy elements would be transferred to a Commissioner who is likely to report to a Home Secretary - the politician most responsible for the use of CCTV by the police or national security agencies.

If such a Commissioner were to be established, it would create a potential conflict of interest as the politician responsible to Parliament for the bodies that use the products of CCTV surveillance will also to be politically responsible for protecting those who are the subject of surveillance. However it has to be recognised that this system of governance is the model that is widely implemented in the UK in relation to Home Office functions whenever policing or national security is involved. Additionally, if there is CCTV specific legislation, such legislation – as with the ID Card Act -  has the potential to allow the Home Secretary to create statutory rules that identify broad statutory purposes for CCTV, specify long retention times for CCTV images, and legitimise wide-ranging, data sharing of images.

In other words, although the Strategy proposes an extensive, integrated, expansion of CCTV surveillance, the checks and balances that protect the public from misuse of CCTV, and the mechanisms that protect privacy have yet to be defined. Although the Strategy's provisional conclusion is that no changes are needed, in practice, who that regulator is, his powers and his ability to provide redress, all assume immense importance if public confidence is to be maintained (assuming the Strategy's expansion plan is implemented).

CCTV in a Surveillance Society

The predominance of CCTV cameras in the UK has featured highly with respect to the Information Commissioner's concerns over the Surveillance Society and to a large extent, the content of the ACPO Strategy and Home Office research serves to reinforce the view that the implementation of CCTV is sometimes unnecessary.

Some senior police officers are of that view also. In May, the Deputy Chief Constable of Hampshire Ian Readhead (who is one of the leading ACPO figures on the use of technology by the police) told the BBC's Politics Show that CCTV was being used in small towns and villages where crime rates were low. He complained that in his force area includes the small town of Stockbridge, where parish councillors have spent £10,000 installing CCTV and questioned whether the relatively low crime levels justified the expense and intrusion. “I'm really concerned about what happens to the product of these cameras, and what comes next?”, he told the programme. “If it's in our villages, are we really moving towards an Orwellian situation where cameras are at every street corner? And I really don't think that's the kind of country that I want to live in”.

As the Information Commissioner is concerned to regulate the uncontrolled development of surveillance, one can expect that when the Commissioner issues guidance on Privacy Impact Assessments (PIA), that a PIA will be an essential pre-requisite for the implementation of a CCTV system. It follows that if a CCTV system fails to deliver its promised results, then a data controller might have to consider its removal. Indeed, the Strategy's procedures should come to the same conclusion: for example, if the case for a CCTV system does not pass muster in relation to the Strategy's procurement or operational standards, then one assumes such a CCTV system would either be uninstalled, or not be procured, or upgraded to the correct specification.

Concluding comment

So what to make of the above? Although there are undoubted CCTV successes, the tenor of the text produced by all the sources described in this analysis suggests all is not well with the current CCTV infrastructure. The ACPO National Strategy claims that national standards will remove all these problems and that new CCTV networks will become a cost-effective tool that performs to public expectations. Whether the Strategy's claim that it will provide for this cost-effective proof for CCTV is, of course, an unknown quantity. All that is known is that many CCTV systems “do not do what it says on the tin” and if this is the case, the question then arises as to why they are installed.

But what of our readers whose employer has installed or thinking of installing a CCTV system? As data protection officer you must get involved in the choice of CCTV technology and make sure that your employer's CCTV system is not one of those that is unfit for purpose? So become involved in CCTV system procurement; we hope this article has convinced you that this is essential. Above all, when Privacy Impact Assessments arrive – try one out on your CCTV system.

Dr. C. N. M. Pounder

December 2007