By Chris Sarno, Michael Hough and Marjorie Bulos
 Criminal Policy Research Unit
South Bank University
August 1999



Background to the schemes
Scheme operation
CCTV controller and police interviews
Codes of practice
Awareness of schemes
Scheme effectiveness
Overview of crime and disorder in Southwark
Impact on crime – overall assessment
Impact on recorded crime in the Elephant and Castle
Impact on recorded crime in Peckham
Impact on recorded crime and disorder in Camberwell
Impact on recorded crime and disorder in East Street
Cost effectiveness of systems
Summary of findings
Good practice lessons
Continuing evaluation
Control room management



We would like to thank a number of people who have helped the research team to undertake this piece of work. Stan Dubeck and Aelswith Frayne were consistently helpful in providing us with contacts and background information about the schemes. We would also like to thank Spencer Chainey, Romy Conroy, Chris Arnold and Sergeants Guy Beattie and Steve Mumford for their help in mapping areas and compiling crime and disorder data. We are very grateful to the London Borough of Southwark for funding the research.

We are additionally grateful to operators and their managers for being so accommodating throughout the duration of the research. Finally, thanks goes to all police officers who agreed to be interviewed and helped to provide data in other ways.

Chris Sarno

Michael Hough

Marjorie Bulos




This report sets out the results of an evaluation of four CCTV schemes in the London Borough of Southwark.

Key findings about the impact on crime

In assessing the impact of CCTV on crime, we examined trends in the four 'target areas'. These have been compared with trends in the 'buffer zones' immediately surrounding the target area, and with 'comparison areas'. Where possible we looked at recorded crime rates in the two-year periods before and after implementation. When we refer to ‘recorded crime’ we have excluded the 20% or so of recorded crimes which could not conceivably be affected by CCTV. Key findings to emerge across the sites are:

  • recorded crime across Southwark fell during the implementation period. In the year ending January 1999 it was 5% lower than during the previous twelve months
  • in the year following installation recorded crime in target areas fell by between 10% and 12% compared to pre-CCTV levels
  • recorded crime in the target areas fell more than in the comparison areas
  • with the exception of Camberwell, reductions in crime across the buffer zones either matched or outstripped the target areas
  • crimes most affected by CCTV were: burglary, criminal damage, street and vehicle crime.

Elephant and Castle

  • there was a 17% fall in recorded crime in the target area in the two years following the introduction of CCTV
  • recorded crime in the buffer zone also fell by 17% in this period. The effect of CCTV may have spread to the buffer zone, or there may simply have been an overall downward trend in the area regardless of CCTV
  • however there was a very steep fall in street crime (robbery and theft from the person) which fell from 154 in the year preceding installation to 59 in Year 2 post-installation, which can be attributed at least in part to the CCTV system
  • assaults and public order offences increased in the post installation period. Some of this increase may have been the result of the Stephen Lawrence Inquiry being held in the area.


  • we have been unable to undertake a full ‘before and after’ analysis of recorded crime data in Peckham. This is because the CRIS system which provided us with crime data for this study was not operational until after CCTV had been installed in Peckham
  • some manually collected police data is available for the ‘before’ and ‘after’ period. Street crime and burglaries within the field of vision of the cameras fell sharply in the two years following the installation of cameras
  • a ‘before-after’ analysis for the beat buffer zone (town centre beat areas) was undertaken using data provided by the Crime Management Unit in Peckham. Data were only available for the months of August and September for the two years preceding and following camera installation. These data showed a 17% reduction in recorded offences between the ‘before’ and ‘after’ periods
  • at least in part the falls can be attributed to CCTV. However the methods used by the police to collect this data were different from our own and do not take into account the possible displacement from one type of crime to another or geographical movement of crime to other adjacent areas.


  • recorded crime in the Camberwell target area decreased both before and after the installation of CCTV. It fell by 4% in the twelve months prior to installation and was down a further 12% in the year following the introduction of cameras
  • there were significant reductions in street crime, vehicle crime and violent crime although these had already been decreasing prior to the introduction of cameras
  • following the introduction of cameras recorded crime in the buffer zone and comparison sector increased by7% and 3% respectively. These increases may be the result of a shift in crime out of the target area or alternatively a natural rise after steep reductions in the previous year’s figures
  • disorder calls to the police increased by 4% in target area during the post installation period. This was somewhat less than the 14% rise over the same period for the buffer zone.

East Street

  • recorded crime in the target area increased by 3% in the twelve months prior to the introduction of cameras. This was followed by a 10% decrease in the year following installation
  • vehicle crime (down 44) and criminal damage (down 23) both decreased substantially. There was a rise in street crime from 156 to 190 offences. This conceals notable variations by crime type. Robberies halved but thefts from the person increased from 106 to 164
  • recorded offences in the buffer and comparison zones decreased at a higher rate than the target area both prior to and after the introduction of cameras. It is difficult to unpick whether this is the result of a diffusion of benefits from the CCTV scheme or part of an overall downward trend in this locality
  • there was a 9% reduction in disorder calls to the police in the target area. There was little change in the buffer zone.


Control room operation

  • few incidents were identified at the Cerise Road control room, thus detections were low
  • questions were raised about reliability of equipment in Camberwell and East Street
  • police said schemes could be used more effectively and they felt partly to blame for the systems' limited success in detecting crime
  • CCTV controllers had little formal training in what to look for and what constitutes suspicious behaviour
  • the codes of practice need review in the light of their operation and the LGIU model code.

Public perceptions of CCTV

  • 55% of those interviewed in the public perception survey believed crime had fallen
  • violent attacks and robberies worried respondents more than other types of crime
  • about 90% of respondents said they felt safe during the day. This fell to about half in the evenings
  • almost two-thirds who knew cameras were present said they felt safer as a result
  • 90% felt CCTV helps to catch criminals. Two-thirds believed cameras deter crime and make the public feel safer

The response of local businesses

  • 33% of representatives of local businesses felt there had been a reduction in crime
  • those running local businesses felt the presence of cameras should be more highly publicised
  • half of them felt more secure at work; half also thought customer confidence in the area had increased
  • local businesses were unwilling to contribute to the cost of CCTV, believing this to be the role of local and central government


The introduction of Closed Circuit Television (CCTV) systems into public spaces has been both rapid and widespread. Research by Bulos and Sarno (1994) for example showed that 75% of schemes operating in 1994 were less than three years old. This growth has been fuelled more recently by the Home Office which has provided financial support believed to be in the region of £50 million over a four-year period.

In Southwark, systems introduced into public spaces were primarily intended to deter and police criminal and anti-social behaviour within a framework of town centre management. This was prompted by local policies emphasising the need to reduce crime and fear of crime as part of the regeneration process. The primary driving force behind CCTV in Southwark, as in many other localities, has been the local authority, who in partnership with the police, local businesses and sections of the community have contributed significantly to the implementation and operation of systems (Bulos and Sarno, 1996:4).

This report presents the results of an evaluation of systems in the Elephant and Castle, Peckham, Camberwell and East Street. The evaluation focuses in particular on:

  • the schemes’ impact on levels of recorded crime
  • effect on fear of crime and consequent behaviour
  • operational and management structures
  • influence upon local businesses
  • technical performance.


The research covered in this report was conducted between June 1998 and May 1999. A range of information was collected: statistics of recorded crime and disorder; data from system logbooks and repair invoices; interviews with the police and CCTV operators; a survey of town centre users; and a survey of local businesses in each of the four sites. This enabled us to measure (with some limitations):

  • impact of CCTV on crime, including any displacement effects
  • influence of CCTV on the fear of crime and on consequent town centre use
  • management and operation of the systems, including codes of practice
  • scheme cost-effectiveness
  • effect of the scheme on local businesses.


Crime and Disorder Data

Recorded crime and disorder statistics were collected with the aid of the Crime Recording Information System (CRIS) and Control and Despatch Management Information System (CADMIS). CRIS was introduced in Southwark in February 1996 and holds crime allegation data. Information on public disorder was collected from CADMIS. Southwark Council and the three police divisions use a software package, Omnidata, which enables CRIS and CADMIS address data to be "cleaned" and geo-coded to a high degree of accuracy [1]. This enables relatively straightforward data collection for specific areas and consequent crime mapping. Data was collected for three principal types of area for each scheme:

  • target areas (those within the field of vision of CCTV cameras);
  • buffer zones (surrounding beat areas) to measure displacement and diffusion of benefits; and
  • comparison areas to provide some benchmark information on trends, to help address the question of cause and effect.

Target areas were defined as areas under direct surveillance within 200 metres of each camera. At this distance on "full zoom" an adult’s image fills approximately a third of the video monitor. We established the boundaries of each target area using this "third of monitor" rule of thumb, panning each camera through its full field of vision. Buffer zones were based on police beats surrounding the target localities. Data provided for these zones exclude figures from target areas. Comparison areas consist of Southwark Metropolitan Police Force sectors[2], divisions and the borough as a whole. In collecting data from target, buffer and comparison areas we believe this approach should overcome any variations which might occur in reporting and recording. Table 1.1 below defines buffer and comparison areas for each site[3]. Maps of each scheme and surrounding areas are in Appendix D.

Table 1.1 - Definitions of target, buffer and comparison areas for each site scheme


Buffer Zone
(excludes figures
from target area)

Comparison Sectors
(excludes figures
from target area)

Comparison Divisions

Elephant and Castle

Walworth Division beats 1, 4 Southwark Division beats 3, 7

Newington and Borough

Walworth (MS) and Southwark (MD)


Peckham Division beats 1, 2 and 4


Peckham (MM)


Walworth Division beats 11, 13 and 15


Walworth (MS)

East Street

Walworth Division beats 3, 5


Walworth (MS)

Recorded offences[4] falling within these areas were collected on a monthly basis from February 1996 to January 1999.[5] Crimes unlikely to be affected by CCTV such as deception, handling stolen goods and malicious phone calls, were excluded. We have also left out drug offences. This is because we feel that policing activity, particularly stop and search, is directly linked with rises in recorded drug crime. For the borough as a whole drug offences rose by 43% between the year ending January 1997 and twelve months to January 1999. This is likely to be the result of more intensive policing of drugs to meet targets rather than a huge increase in drug use across the borough. The main categories of crime data collected included:

  • Non residential burglary (EB, ED)[6]
  • Residential burglary (EA, EC)
  • Criminal damage (CY, BF)
  • Theft from vehicle (VE)
  • Criminal damage to vehicle (VD)
  • Vehicle interference (VF)
  • Theft of motor vehicle (VA, VG)
  • Robbery (CE)
  • Theft from the person (CG)
  • Serious assault (CA, CB, CC)
  • Section 4 Public Order (BE)
  • Assault-ABH (BG)
  • Sexual assaults (CR, CS, CT, CU)
  • Shoplifting (BB)
  • Cycle thefts and other thefts (BA, BD)
  • Other assaults - common assault and assault on a police officer (BH, BK)
  • Racial incidents (BR).


These crimes constitute about 80% of the total number of offences recorded in Southwark in the period April 96 – March 98. Throughout the report we have aggregated these categories into six main groups as follows:

Burglary: Residential burglary; Aggravated residential burglary; Non-residential burglary; Aggravated non-residential burglary

All Vehicle Crime: Theft from motor vehicle; Theft of motor vehicle;Taking conveyance; Arson to motor vehicle; criminal damage to motor vehicle; vehicle interference

Street Crime: Robbery; Theft from the person

Sex/Violence: Rape; Indecent assaults (male and female); Indecent exposure; Section 18 assault; Section 20 assault; Murder; ABH; Common assault; Assault on police officer; Section 4 – public order offences

Theft: Theft of pedal cycle; Shoplifting; Theft – other

Vandalism: Criminal damage (over £5000); Criminal damage (under £5000)


Table 1.2 sets out the pre- and post-installation periods. In the case of the Elephant and Castle (implemented January 1997) this enabled the year prior to and two years following installation to be measured. For the Camberwell and East Street we examined the two years prior to installation and one year following. However, in Peckham CCTV was installed about 4 months prior to CRIS going live in Southwark. This caused a considerable problem in collecting crime data for the ‘before’ period and four months following installation. A limited amount of information on street crime and burglaries for the pre- and post-installation period is available for those areas under direct camera surveillance. We also have, through the Crime Management Unit in Peckham, assembled further data enabling limited examination of the changes for the town centre beat areas. Adequate data is available to examine the months of August and September for the two years preceding and following implementation. Although beats do not correspond directly with target areas (being much larger) was the only feasible method available.

Table 1.2 - Installation dates for systems and periods examined


Month CCTV installed

Year -2


Year -1


Year 1


Year 2


Year 3


Peckham ( street crime and burglary)

Oct 95

Oct 94–Sep 95

Oct 95–Sep 96

Oct 96–Sep 97

Peckham ( Town Centre beats data)

Oct 95

Aug 94-Sep 94

Aug 95–Sep 95

Aug 96–Sep 96

Aug 97–Sep 97


Peckham (CRIS data)

Oct 95



Feb 96 – Jan 97

Feb 97 – Jan 98

Feb 98 –Jan 99

Elephant and Castle

Jan 97


Feb 96 –Jan 97

Feb 97 – Jan 98

Feb 98 – Jan 99



Jan 98

Feb 96 – Jan 97

Feb 97-Jan 98

Feb 98 – Jan 99



East Street

Jan 98

Feb 96 – Jan 97

Feb 97–Jan 98

Feb 98 – Jan 99



Public Perception Surveys

A quota sample of two hundred individuals was interviewed in each site using a structured questionnaire.[7] Quotas were set to reflect the age, sex and ethnic background of residents in all wards covering each target area based on data from the 1991 Census. The questionnaire took about five minutes to complete, asking questions on respondents’:

  • knowledge of the scheme
  • whether he/she believe it was effective
  • how cameras affected the way he/she felt about and used the area


The surveys were undertaken on different days of the week and times to reflect the different patterns of use at each site. The lack of precise baseline data for the period prior to the introduction of cameras meant a modest snapshot view of opinions was the only realistic approach. People were asked to comment retrospectively on changes to the area.

Interviews with CCTV staff and the police

Interviews used a structured questionnaire comprising a mix of qualitative and quantitative questions. Data from the interviews were coupled with observational work and the interrogation of documentary materials to construct a picture of the management, work practices and technical performance of systems. CCTV operators were questioned on their experience of using the system; operational matters including tape management; control room procedures and management; problems with the system and possible improvements; perceived impact of the system on crime; and the codes of practice. We interviewed three operators at each control room, and eight police officers, including home beat officers, crime prevention officers and sector inspectors. We sought to establish how they believed the system was being used and for what purpose; changes in policing tactics; relations with control room staff; quality of evidence produced; and the codes of practice.

Business Survey

A small-scale postal survey was carried out of local businesses in each of the four sites. A random selection of 150 premises was drawn from lists of local businesses provided by the council. The questionnaires were designed to collect views on: crime and safety in the area and the impact cameras have made; perceived impact of cameras on businesses; and future funding arrangements of CCTV systems. Each questionnaire was sent out with a covering letter explaining the purpose of the research and a business reply envelope. To boost the response rate questionnaires returned within 5 weeks were entered in a prize draw for £50.

The initial response rate from the postal survey was poor: of 600 questionnaires only 46 were returned. Given this it was decided in agreement with Southwark Council that the sample would be boosted by door-stepping local businesses. Fieldworkers undertaking the public perception surveys called on every second business in each of the four areas. The minimum quota for each site was 30 completed questionnaires. Market stalls in East Street and the Elephant were included. In total 134 questionnaires were completed: 44 in Camberwell; 34 in East Street; 31 in Peckham and 25 in Elephant and Castle.

Logbooks and repair invoices

Data was extracted from CCTV repair logbooks for four months from June to September 1998. This was to verify information from interviews on: evidence gathered from the system; its use by the police; and equipment reliability. The following were analysed:

  • number of tapes viewed by the police and the outcome
  • number of times the police took control of the cameras and for what duration
  • number of equipment failures.

Problems of measurement and causality

In evaluations of this sort there are few reliable measures of crime and anti-social behaviour. Accurate measurement of actual levels of crime can only be done at very great expense through complex crime surveys.[8] Recorded offences, which are used in this report, only form a proportion of the total crime because many crimes go unreported and unrecorded.[9] Small changes in reporting or recording practice can lead to changes independently of the underlying crime rate. However, we have no reason to believe these varied within the borough and thus the results are unlikely to be seriously affected by such changes.

Even assuming that crime and disorder statistics faithfully reflect the underlying trends, there are problems in identifying what has caused the change. Levels of recorded crime change over time. During the period being examined, recorded crime was falling across the borough (see below). It is difficult to distinguish the impact of the four CCTV schemes from other contributory factors of which there are many. The two most important include the regeneration of Peckham town centre and the "five estates" area and the Met-wide Operation Eagle Eye. Both may have independently affected crime rates.

The Peckham Partnership secured a £60 million grant from the Single Regeneration Budget programme over seven years to regenerate the town centre and five estates just to the north of the High Street. The strategic objectives include:

  • bringing jobs to the five estates residents and prosperity to Peckham
  • providing young people with basic and flexible skills
  • transforming the five estates into a desirable residential area
  • making Peckham an area where people feel safe
  • giving Peckham a vibrant and viable town centre
  • improving physical and mental health
  • providing safe and convenient access for people.


This initiative has involved a number of changes to the layout of the town centre. As a result of these changes cameras have been moved and new areas are now under camera surveillance. The demolition and rebuilding of the Sumner Estate and construction of a new Health and Leisure Centre in the Canal Head locality has resulted in new residential roads, one of which, Jocelyn Street, is now under camera surveillance.

Operation Eagle Eye was a response to the increasing concern over the rise in street robberies in the early 1990s.The Metropolitan Police Service (MPS) established a steering group in 1993 to prepare a strategy for tackling street robbery and launched Eagle Eye two years later in October 1995. It covered 25 divisions with high robbery rates, including Peckham and Walworth. The aim of the operation was to improve performance against street robberies and increase the detection rate to 15% during 1995/96. This was to be done working in partnership with the local community showing high levels of ethical standards and care for the victims. Across all Eagle Eye divisions there was a 5% drop in robberies during the first year of operation.

The main solution to these problems of measurement and causal attribution is to make a detailed comparative examination of trends in target areas, buffer areas and comparison areas. If the target area shows a greater reduction than buffer or comparison areas, and if the net reduction in the target and buffer area is greater than in the comparison area, this provides evidence of impact. Any changes in recording or reporting are likely to have affected all areas equally.

The remainder of this report presents the detail of the findings. Chapter 2 presents findings relating to control room operation, repairs, visits by the police and interviews with CCTV staff and the police. Chapter 3 considers public perceptions and views of local business communities. Chapter 4 sets out our analysis of the impact of the four systems on crime and disorder. The final chapter presents an overview of the results and discusses their implications.



Several factors determine the extent to which CCTV can impact on crime. The ‘micro-geography’ of the area is likely to interact with the design of the system, its general visibility and the extent of coverage to achieve greater or lesser effect. Furthermore, the situation in which CCTV is placed may affect how and whether it ‘works’. Ditton et al (1999) note that CCTV may ‘work’ better in a small town where local police and residents know faces than a large city where this is less likely to be the case. The way in which the system is implemented may be of equal importance. The deployment of signs, publicity advertising the system’s introduction, and strategically placed reports of the system’s success could - depending on handling - have a powerful deterrent effect in their own right, at least in the short term (Tilley 1993).

To be effective the systems themselves should be visible as Tilley notes (1993:332)

‘To gain maximum effect the CCTV must be overt. Thieves and vandals do not want to be caught, so make sure they know you have an effective system in operation.’

Equally, Brown (1995:vi) identifies the nature of layout and degree of camera coverage as significant factors that may impact on levels of crime. He draws attention to the Newcastle system as being highly effective because:

‘ ... the layout of the town centre is simple and the degree of camera coverage is high’.

The success of CCTV in reducing crime in car parks provides a clear example of the significance of location, type of crime and design. Maximum impact seems to result when the following factors coexist:

  • defined spaces with a clearly specified use
  • low levels of activity
  • narrow range of activities
  • a layout which makes for maximum camera coverage
  • a form of crime which is often opportunistic
  • easy access by enforcement agencies.

This chapter considers how far the four schemes covered in this report exhibited these factors. It discusses how and why the different schemes were established; who was involved; what the schemes looked like; and examines the views of those involved with the schemes.

Background to the schemes

The Elephant and Castle

The area is dominated by a large shopping centre surrounded by a network of arterial roads inter-linked by roundabouts. Access to the shopping centre on foot is primarily through a system of complicated subways, which prior to the installation of cameras had achieved some notoriety as robbery hotspots. What the area lacks in charm it makes up for in transport links with an underground and railway station and innumerable bus stops. Main roads leading away from the Elephant are lined with a mix of commercial premises and public sector housing. Large employers in the area include South Bank University, the London College of Printing and the Department of Health.

In 1993 a community safety committee composed of representatives from the council, police and local businesses was set up to examine ways of reducing crime and the fear of crime in the Elephant and Castle. Police operations had been successful in reducing crime in the area, but senior officers regarded these as expensive and only a short-term solution. Committee members believed that a 24-hour surveillance system was the most effective means of dealing with the area’s problems.

In January 1997 thirty-four cameras were introduced into the Elephant and Castle covering the subways, bus stops and streets surrounding the shopping centre. A map of the area can be found at Appendix D. The system, along with fifteen cameras in the shopping centre, is monitored from inside the shopping centre and linked to Walworth Police Station. Six external and twelve internal cameras can pan-tilt-and-zoom (PTZ). The capital cost for the scheme was £310,000. Two-thirds of the total cost were provided by the Home Office and Southwark Council. Running costs for the scheme are shared between UK Land Limited, who operate the shopping centre, and the Council. The main aims of the scheme are to:

  • reduce opportunist street and subway crime
  • eliminate drug trafficking in the area
  • increase confidence amongst members of the public using the area after dark.


Closed Circuit Television was introduced into Peckham Town Centre in October 1995. During the early 1990s the town centre had shown clear signs of decline, with a number of major chain stores including Marks and Spencer and British Home Stores closing. In the years leading up to the introduction of the scheme, the area had become a hotspot in the borough for street robberies and drug dealing. Police statistics for 1994 indicated that in the High Street and Rye Lane alone 310 robberies were reported - 16% of the borough’s total for that year. The system was designed to address these problems, with the following main aims:

  • deter and prevent crime, particularly street robbery and drug dealing
  • reduce the fear of crime and provide reassurance to the public
  • assist the police to intervene in the commission of crimes
  • provide prosecution evidence in the event of crimes
  • reduce the incidence of vandalism, graffiti and other criminal damage
  • prevent and provide a response to racial harassment
  • facilitate traffic management.


The on-street system, which comprises fourteen pan-tilt-and-zoom cameras, was added to the pre-existing 27 static cameras in the Cerise Road car park. Developed on a partnership basis between the council, police, local traders and the community, the system was installed as part of a wider regeneration strategy. The local authority met the bulk of the £366,000 capital cost of the scheme. Cameras were primarily located at junctions along Rye Lane and the High Street giving coverage of a number of adjacent side roads. A map of the target area can be found at Appendix D. Both schemes are monitored 24 hours from a control room at the Cerise Road car park. The main town system is linked to Peckham Police Station.

The area under camera surveillance is primarily commercial, dominated by a mixture of small local businesses and a number of larger high street retail chains primarily located in the Aylesham Shopping Centre. Dwellings along the High Street and Rye Lane constitute a small number of flats above shops. Adjacent roads, however, particularly south of Peckham Rye Rail Station (i.e. Choumert Road and Blenheim Grove) are more residential with a high proportion of terraced and semi-detached housing. The High Street forms a main arterial route (A2) into Central London from the east. Rye Lane is a significant thoroughfare for public transport to and from the south of the borough.


Following a successful bid by a council-led partnership to the Home Office, cameras were introduced into Camberwell town centre in January 1998. Prior to the introduction of CCTV, Camberwell had suffered from high levels of street crime. Between September 1995 and August 1996 police recorded 368 robberies in Camberwell – 85% of which happened in the three-town centre beat areas. As in the Elephant and Castle, short-term initiatives were undertaken by the police to combat robberies during peak times including Christmas, Easter and the summer holidays. Whilst this had the effect of reducing crime over the short-term it was generally regarded as an inefficient use of police resources. In the light of this a partnership was established to examine ways of implementing a CCTV system. The main aims of the system are to:

  • reduce street crime
  • identify shoplifters and robbers operating in the area
  • increase the viability and confidence of small independent traders
  • safeguard all members of the public and consequently reduce the fear of crime
  • provide a system comparable to that in Peckham
  • supplement police resources.

Camberwell’s system consists of 17 pan-tilt-and-zoom cameras. The scheme was designed to cover the main arterial routes running through the commercial heart of Camberwell including Denmark Hill, Camberwell Church Street and Camberwell New Road. These main roads are commercial in nature consisting primarily of small locally owned shops and restaurants. Residential side roads including Bessemer Road, De Crespigny Road and Grove Lane are also well covered. The system is reckoned to include 17 roads, in total covering 250 commercial premises and 5,000 square metres of pedestrian routes. A detailed map of the target area can be found in Appendix D. The scheme cost £307,000 to implement and is monitored from the Peckham control room at Cerise Road. An additional monitor is located at Walworth Police Station.

East Street

East Street is the site of one of London’s oldest surviving street markets. The market site runs off the Walworth Road half a mile south of the Elephant and Castle. In recent years, the area, which consists of a mix of commercial and residential premises, has been in decline and suffered from increasing levels of crime.

A partnership initiative between the council, police and local traders was established with the goal of regenerating the area. The East Street system started operation in January 1998. The scheme, consisting of 11 pan-tilt-and-zoom and 1 fixed camera, was intentionally located in the market area to:

  • deter and detect crime
  • reduce the fear of crime
  • provide evidential quality footage of incidents and perpetrators
  • restore confidence in the area as a viable commercial location.

The area under surveillance extends a third of a mile to the east of Walworth Road and covers 8 adjacent streets and a number of local free car parks which in recent years have been hotspots for car crime. A detailed map of the area is located in Appendix D. Set up at a cost of £168,000 the scheme is monitored from the Cerise Road control room and linked up to Walworth police station. The costs of the system were primarily met through local traders and the Home Office.

Scheme Operation

This section provides an overview of the operation of each scheme based on data collected from control rooms, interviews with police and CCTV controllers and direct observation. The schemes operate in different ways, reflecting variations in the way each scheme is funded, operated and owned.

Each control room in Southwark has adopted a common CCTV code of practice. This code is designed to govern the management of the council’s control rooms and facilities. The code covers:

  • installation of CCTV (i.e. consultation and privacy)
  • tape management (i.e. loading ,access, use, reuse, storage, and cataloguing)
  • maintenance of an incident book
  • procedures for the police
  • control room management including access, transfer of monitoring, special contingencies, and communications
  • provision of stills
  • monitoring and review of the code of practice.


As part of control room monitoring an independent panel of 10 people from the local community has been established. Nominated by the Police Community Consultative Group, the panel are authorised to undertake spot checks to ensure systems are operating in accordance with the code of practice. Later in this section we shall compare the code to the Local Government Information Unit’s model code.

Scheme operation at the Elephant and Castle control room

Burns, a private security company, operate the Elephant and Castle camera system. Managed by UK Land Limited and funded in association with the council this system is monitored from the Elephant and Castle shopping centre. The control room is multi-functional acting as both a monitoring site and a reception area for delivery vehicles to the shopping centre. Two members of staff operate cameras during each shift. An additional member of staff deals solely with delivery vehicles. Operators work twelve-hour shifts.

During the period June to September 1998 the police made 161 visits to the Elephant control room. This figure is inflated due to the Lawrence Enquiry. Sixty-eight tapes were taken for examination during this period (31 between June 29th and July 9th for the Enquiry): 59 by the Walworth Division; 6 by Customs and Excise; and 3 by the Southwark Division.

Table 2.1 Tapes taken by police from the Elephant control room (6/98-9/98)

Month (1998)

Number seized











There is no record of how often police viewed tapes at the control room. Control room staff often view tapes where an incident has occurred in the shopping centre on behalf of police officers to save police time in finding the relevant tapes and then reviewing them[10].

It is worth emphasising several features of the Elephant and Castle system which were absent from the other three:

  • There are many sources of information on the ground particularly within the shopping centre. These include: centre security guards; shop security and staff; and members of the public who can report incidents to security staff or the shopping centre information booth. As one CCTV controller noted, ‘Often we get a call off the guards on the shop floor that someone looks suspicious – so we watch them.’
  • The police have established an ‘Operation Watch’ scheme in the Elephant and Castle. This consists of a picture file of 10 individuals held at the control room who are believed to be committing offences in the area. The file is updated quarterly by the intelligence unit at Walworth Police Station. Whenever an individual in the file is seen in the Elephant details of their clothing and movements are recorded and passed on to the police. Also the control room is visited regularly by the local home beat officer. Intelligence information and feedback is passed between the officer and control room staff. One officer suggested, ‘Over the last year I have tried to make the control room here part of our police station because there are so many benefits and we have such a close working relationship.’ This appears to have fostered a proactive relationship between staff and the police.
  • A number of controllers have worked as security guards in the shopping centre for many years. The three controllers interviewed had 15 years’ experience of working in the shopping centre between them. They are therefore well placed to recognise individuals who may have committed crime in the area in the recent past. This was noted by two of the CCTV controllers, one of whom maintained, ‘Over the last three years I have got to know the area - know the people. I know the criminal ones. I spot them and watch them.’
  • The layout of the area and number of cameras installed (in and around the shopping centre) makes for very good camera coverage. The subways are the main routes for pedestrians moving in and around the Elephant and Castle and are extremely well covered. The possibility of getting good quality close-ups of possible suspects is high. Although obstacles including buses, trees, market stall and advertising hoardings are a problem, the viewing of pedestrian routes in all other sites is more limited.
  • Operators regularly track individuals who are seen to be acting suspiciously or are known to the police and operators as repeat offenders. When successful, this process can pre-empt incidents and enable rapid and appropriate responses. What makes this type of system efficient is that controllers have direct contact with shopping centre security staff facilitating quick and direct action when suspicious behaviour is detected or an incident spotted.

Despite the relative success of the system in preventing and detecting crime a number of bad practices and discrepancies in logbooks were noted during the course of the research. Whilst we do not believe that it would be constructive to highlight particular instances observed by the research team current structures need review and possible amendment Particular examples of practice at odds with the code are:

  • viewing tapes without the presence of the police officers
  • lack of a logbook showing when police viewed tapes
  • tapes not always being signed out by police officers.

To ensure adherence to the code of practice this control room would benefit from tighter management and regular and extensive spot checks. We believe that there are possible civil liberties implications over maintaining police photos of known and possible offenders at a control room outside of a police station. Recommendations relating to control room practices are set out at the end of this report.


Scheme Operation at Cerise Road control room

The Peckham, Camberwell and East Street systems are all operated from a single control room in Cerise Road, Peckham.[11] Managed by the council and operated by Bolens Security, the Peckham, Camberwell and East Street systems were established primarily to deter and reduce crime. As with the Elephant scheme cameras are monitored for 24 hours a day. Three controllers, working twelve-hour shifts are on duty at any one time.[12]

Although detection and apprehension are key aims of the system, local police officers involved in establishing the Peckham scheme suggest: ‘It was put in to reduce and deter rather than to arrest. If it had been done to do that, I think it would be a failure.’

Evidence gathered to date seems to support this statement. Operators catch few incidents on camera. Between June and September 1998 the police viewed tapes 59 times. On 18 occasions incidents were recorded on the tapes. Only 8 original tapes were taken during that period. More recently between January and April 1999 forty-one tapes were taken. Nevertheless, as one controller maintained ‘the chances of viewing an incident on camera are extremely rare.’ When arrests occur this is usually subsequent to the incident where pre-set camera has recorded up an offence. Probable explanations for this are discussed below in more detail.

Operators spend much of their time scanning target areas looking for suspicious behaviour and incidents. This is their main means of gathering information and evidence. Although relations between the controllers and police appear to be good there is scope for further co-operation. Little direct intelligence information appears to be passed to operators from the police. Compounding this, direct radio links between security staff in the Aylesham Centre and CCTV operators are used infrequently. [13]

The police frequently take control of the system, particularly in the evening. In the four-month period from June to September 1998 the police took control of the cameras on 181 occasions, amounting to a total of 200 hours of control. Incidents reported to the CAD room at Peckham Police Station from the public or officers on the ground regularly lead to police use of the system.

CCTV controller and police interviews


Although all controllers interviewed had some general training as security guards, none had had any intensive formal training either in operating CCTV systems or in the principles of crime prevention through surveillance. Inevitably they had acquired experience on the job; and they had their own views about offenders’ methods of operation and on what constituted suspicious behaviour. One controller said, ‘Nobody has told us what to look for. You just pick up on things. Like I said I have been working here for 8 years. So you know some of the locals. You get the new faces - sometimes you just get characters that just look dodgy. They’ve got a big bag on them and they don’t look like they’re going shopping’ Another saw suspicious behaviour as one person closely following another or an individual hanging out in an area for a long time but not doing anything in particular. A third controller maintained, ‘All thieves don’t work the same. Some of them will walk straight into the shop and come straight back out again. Others will hang around looking left and right and make sure there is no security around and then walk in.’

Although police officers were confident about the abilities of controllers, they felt that they might benefit from some training from the police about the way offenders operated and things to look for. The police have already in some instances given training to controllers on witness statements and court hearings.

Control room practices

Controllers were all well versed about operating procedures set down in the code of practice. Procedures for tape use, storage and cataloguing were all well covered. Knowledge on what to do during an incident was acceptable as was security at the monitoring centres. At Cerise Road the incident book was well maintained - as was the log for police viewing and taking tapes. At the Elephant no records of tapes viewed by officers on-site was kept. Controllers generally accepted that a monitoring panel was a good idea, and all acknowledged the need for codes of practice. For example:

‘… so that people are following the rules and not doing things they shouldn’t be doing.’

‘ ... it’s a set of rules to be followed. We have to know procedures of what to do.’



Controllers at the Elephant and Castle generally felt equipment was reliable. Between June and September 1998, the company responsible for maintaining equipment, Orbis, made twelve visits to the Elephant control room to undertake nine repairs, the majority of which were to fix cameras. At the Cerise Road controllers were happy with the Peckham system. There were however a number of concerns over the Camberwell and East Street systems. Complaints included frequent equipment failures and poor quality images in areas of low lighting. Basic testing of the Camberwell system highlighted some of these problems. When tested two of the cameras were not operational. One controller felt ‘there hasn’t been a day when those systems have been working to their full potential.’ Thirty-seven faults were reported between June and September.[14] There were 59 visits to undertake maintenance work. The most frequent call outs were for camera faults (20). The warranty on the Camberwell and East Street system has been extended. This is because the systems are operating at a level some way short of what is required.

Controllers in the Elephant felt non-rotation or focussing of cameras (particularly after heavy rain) and the loss of pictures were the most common faults. At Cerise Road the main complaint was about fibre-optic links failing. There were some complaints from controllers at Cerise Road about the capabilities of cameras in East Street. One controller claimed the cameras were ‘Very, very slow. They’re very poor. But that’s what the traders wanted, that’s what they paid for.’ Equipment testing (in the presence of a researcher) showed this to be the case when compared with other systems. Generally speaking controllers felt repairs were undertaken quickly and effectively – usually within 48 hours. Some instances were however noted by both the police and controllers where cameras were out of operation for weeks at a time.

Quality of images produced and detections

Controllers also felt that the systems provided good quality pictures that were of evidential quality. There was a general recognition that the quality of images was reduced when it was very sunny, windy or raining heavily. The police were less positive about the capability of the systems. One officer stated, ‘ If cameras are monitoring an incident and focused on a suspect then the clarity is good – but straight recording [i.e. camera in pre-set position on time lapse] is poor quality.’ Two other officers felt that the quality of images produced in the evening in Camberwell was not as good as it could be. Officers interviewed in all sites noted that systems had not met their potential in terms of crime detection. ‘I would like to see the cameras moving about a lot more. If they are stationary they lose their effect.’

Crime reduction and displacement

Controllers and police officers thought that the introduction of cameras had reduced recorded crime. Street crime (in particular robberies) was reckoned to be affected the most. Nevertheless, it was also felt that crime had been displaced. One controller believed, ‘If you’re a regular thief or shoplifter you’re not going to give up your job. Obviously you ain’t going to come into this area – so you’re going to go somewhere else.’ In the Camberwell target area police officers said they had noticed functional displacement of thefts from the person from on street (particularly bus stops) into cafes and pubs. Similarly in Peckham officers commented that drug dealing had been shifted either indoors or into side streets. The police have reacted to this by mounting special operations to deal with these shifting crime patterns.

System effectiveness

Both controllers and officers at all sites - but particularly in Peckham - noted that the system could be used more effectively. Officers believed that they were in part to blame for the systems being unsuccessful in detecting crime. They recognised they could be doing more to help the system work for them. Three officers felt that the systems might benefit from controllers getting police training and greater access to intelligence information. Furthermore, the usefulness of a business and operations watch schemes similar to those running in the Elephant were understood. Police also felt the controllers could ring through to the CAD room more often when the saw something suspicious. Officers at both sites believed that getting controllers more involved in active policing, would make their jobs more interesting, thus leading to the more effective use of the system.

Whilst such activities might improve effectiveness of system operation and are consistent with the recommendations of previous research certain they conflict with the Southwark codes of practice. The code maintains: ‘Whilst engaged in normal surveillance, operators will not linger on members of the public engaged in legal but personal or intimate pastimes.’ Clearly there are quite difficult civil liberties issues here which would need to be incorporated in any discussion about improved efficiency of the system.


Codes of practice

Issues around the code of practice and adherence to them have been discussed earlier in this chapter. This section will assess the codes against the Local Government Information Unit’s model code as set out in ‘A Watching Brief’ (LGIU 1996).


The Southwark Code

‘A Watching Brief’ sets out a recommended model code. This is intended as a template from which system owners can construct their own code taking into account local circumstances. The first step in evaluating the Southwark code was to match it against the sections recommended in the model code. The Southwark code of practice appears to be still in draft form. No date is given for its introduction. The document itself has not been signed although organisational entities are specified. It is not clear to whom any enquiries about the code should be made.

The comparison with the model code shows that some matters have been dealt with in some detail, particularly those which relate to detailed day-to-day operational matters and procedures. Others have not. For example, it focuses in detail on the needs of the criminal justice system, but fails to address questions of accountability and information quality (See Table 2.2). Furthermore, it fails to specify that it is used to achieve those objectives originally set for the system which go beyond a narrow interpretation of crime prevention.

Table 2.2 Analysis of Southwark Code of Practice against the LGIU Model Code


Model Code

Southwark Code

a = yes x = no

Purpose Statement



Data Protection Implications

Responsibilities of the Owner




Management of System





Public Information

Residential Areas

Assessment of the Scheme



Breaches of code

Control and operation of cameras

Access and security of monitors

Tapes and recorded materials

Dealing with incidents

Police contact and use



























The objectives for the system and limitations on use are clearly stated. There is an implied hierarchy of use.






Sets out the conditions and control of police and other access and use.





Code review annually (17) but not a procedure for system change.





Not details under specific head. Although now being undertaken



Set out in procedural guide



Whilst there was some variation in the reasons the four CCTV systems in Southwark were installed, it is clear that all the schemes were implemented mainly to reduce the fear of crime and to deter offending. This chapter will examine the results of the public and business surveys undertaken in each site in relation to the first of these objectives whilst considering whether respondents knew about CCTV schemes and how they viewed them. Opinions were also sought from local businesses on their expectations of CCTV; how schemes might be funded in the future and whether they believed cameras had any impact on aspects of their business. The numbers of interviews carried out at each site are shown in Table 3.1. Views from both sets of surveys have been compiled under the following headings:

    • awareness of schemes
    • fears
    • expectations
    • scheme effectiveness

Table 3.1 Number of interviews undertaken in each site


Public Perception


Elephant and Castle









East Street



Awareness of schemes

Across the four sites 67% of respondents in the public perception survey said that they were aware of cameras in operation. Awareness was highest in the Elephant and Castle (74%) and Camberwell (73%) and lowest in Peckham (54%). Ninety-five percent of all representatives from local businesses were aware of cameras. Again awareness was highest in the Elephant and Castle and Camberwell. This could reflect several things:

  • The greater salience of the cameras (especially in subways in the Elephant and Castle, where they are obvious)
  • Better signage
  • Respondents in Peckham may have included a higher proportion of infrequent visitors.[15]

The public were aware of schemes (80% or 548) largely because they had actually seen cameras. A further 11% (75) said they had become aware of the cameras after seeing signage. The most frequently cited areas where cameras were believed to be operating included streets (56% or 321), shopping centres (27%) and shops (24%). In the Elephant and Castle the most frequent answer was the subways (71%).


Violent attacks and street robbery worried respondents more than any other type of crime. This tallies with views across Southwark, where residents considered muggings and violent crimes to be the main problem affecting neighbourhoods in the borough (MORI Southwark Residents Survey, 1998).

Table 3.2 - Crimes which most worried respondents (n = 606)

Crime Type Number of responses % of responses
Violent Attack






Theft of/from vehicle



Theft from bag/pickpocketing



Drug dealing/use



Both men and women were most worried about violent attacks. Females worried more about thefts from the person – 8% compared with 4% (i.e. pickpocketing/bag snatches). Males worried significantly more than females about drug dealing and use (6% compared with 1%). Respondents in the Elephant and Castle were much more inclined to be worried about street robberies (46%) compared with Camberwell (6%). Thefts of and from vehicles caused more concern amongst those interviewed in Peckham and East Street (about 16% of responses in each site). There was little variation by age or ethnicity. Perhaps not surprisingly, however, a greater than average proportion (93% compared with a sample average of 80%) of over 65s were most worried about violent attacks, robberies and thefts from the person.

There were large differences between how the public rated sites in terms of safety during the day and in the evening. During the day 87% of respondents suggested they felt very or quite safe. Twenty-one percent said they felt very safe. More respondents in East Street suggested they felt very safe (48% against 20% across all sites). Nine out of ten 16-24 year olds felt safe compared to only eight in ten over 65s. There were few gender differences. In the evenings these feelings of safety reduced. Less than half, on the other hand, said they felt safe after dark. Only 55% of those who felt safe during the day said they did so after dark. Feelings of insecurity were greatest in the Elephant and Castle. Sixty-four percent said they felt either ‘not too safe’ or ‘not safe at all’ – Only a third said they did not feel safe in Camberwell in the evening. Females were more likely to feel unsafe than males in the evening (55% against 47%).

There were notable differences between how representatives of businesses rated sites in terms of safety and the perceptions of the general public. The business community was less inclined than the public to feel safe both during the day and evenings. During the day only seven in ten (96) respondents said they felt safe - 7% (7) of whom maintained they felt very safe. Although East Street had the highest number of members of the public who felt safe – only a quarter of businesses representatives felt safe during the day.

At night feeling of safety dropped significantly to four in ten (55) respondents from businesses (compared with five in ten of the public). Most worrying of all is that 32 respondent said they felt ‘not safe at all’. East Street again fared the worst with seven in ten from business community (24) maintaining they did not feel safe.

Expectations of CCTV

Members of the business community were asked about their expectations of CCTV. A quarter (32) had no clear expectations. Ten did not answer the question. The remaining 92 respondents gave 159 responses. These were classified into 5 separate categories: 77 felt it would reduce or deter crime; 24 believed they would feel more safe/secure; 12 felt it might lead to an increase in customer trade; and 8 thought it would apprehend criminals. Fifteen offered a range of other responses, for example: system successes would be highly publicised; more cameras would be installed; the cameras should be fully operational and monitored 24 hours a day; and businesses would receive more information on the impact of the cameras on crime.

Respondents were then asked if CCTV had lived up to their expectations. Fifty-one believed it had; 27 said it had not and 21 were unsure. Two-in-five (30/77) of those who expected a reduction in crime stated their expectations had been met. Close to two-thirds (15/24) who believed the area would be more safe/secure felt this was the case.

Scheme effectiveness

Impact on crime and offenders

Both the public and business community were asked about scheme effectiveness. Fifty-five percent of those questioned in the public perception survey felt the presence of cameras had reduced levels of crime; of whom nearly a third believed it had reduced crime significantly. Eleven percent said it had no impact. A third did not know.

Levels of belief in scheme effectiveness were highest in the Elephant and Castle where 66% believed the system had reduced crime. This compares with 36% for East Street (71) and 55% for Peckham. Across these two sites about 40% of respondents did not know. Younger respondents were generally a little more sceptical about the effectiveness of CCTV. Roughly half of 16-24 year olds felt CCTV had any impact on crime. Forty-five to fifty-four year olds were most positive with 62% feeling the introduction of cameras had reduced crime.

Two thirds of business respondents felt that the introduction of cameras coincided with a fall in crime. Only 14 however believed the reduction had been significant. Fourteen suggested there had been no change and two felt crime levels had increased. Responses varied significantly by site. In East Street 14 of 34 respondents felt there had been a reduction compared with 20 out of 25 in the Elephant and Castle.

Crimes which individuals in the street survey worried about most were also viewed as the most likely to be those impacted upon by CCTV. Violence against the person (286), thefts from the person (238) and street robbery (229) were the three most cited responses. Respondents overwhelmingly felt CCTV helps catch criminals; two-thirds also believed the cameras deter offenders from committing crime and makes the public feel safer. The idea that CCTV attracts more people to an area was viewed more sceptically - less half of respondents believed it did so.


Table 3.3 – Public Perception and Business Surveys

Positive perceptions of the effectiveness of CCTV


% agree


% agree


CCTV helps apprehend criminals



CCTV deters criminals from committing crimes



CCTV makes the public feel safer



CCTV attracts more people to use this area



Total number of respondents



Similarly representatives of local businesses also felt CCTV helps catch criminals and makes the public feel safer. Only 53% felt in deterred offenders from committing crime – this was thirteen percent lower than in the public perception survey (66%). Slightly less than half felt it would attract more people to use the area. Traders’ views of which crimes were most common in each site were significantly different from the survey of the public. Whilst violent attack was the most common answer in the public perception survey (48%) theft from the person was the most shared response amongst traders. Street robbery was second – eliciting a response from just under half the sample.

Table 3.4 – Business Survey - Which crimes happen most often? (n = 131)


% of respondents

Theft from person (bag/pickpocketing)


Street robbery


Theft of/from vehicle


Drug dealing/use




Violent attack on a person


Changes in feeling of safety

Perceived reductions in crime fed through into greater feelings of safety. Sixty-three percent of the public who were aware of cameras said they felt safer as a result. About a quarter claimed they felt no different. There was significant variation by site, as Figure 3.1 shows.

Figure 3.1 – How awareness of CCTV cameras reduces fear of crime: survey of general public (n = 635)


Fifty-three percent of respondents in East Street said they felt safer compared with 69% in Camberwell. Half of all 16-24 year olds said they felt safer, which was the lowest of all age groupings; 80% of 35-44s felt safer, which was the highest. Two-thirds of females felt safer compared with 59% of males. Those who did not know there were cameras present were asked if they are likely to feel safer in the future. Sixty-six percent believed they would feel safer.

Six in ten business respondents also felt safer as a result of cameras being installed. Fifty-three said they felt no different and three said they felt less secure as a result. Feelings of safety differed with location. East Street again scored poorly against the other sites with only 13 of 34 individuals mentioning they felt safer. In Peckham the figure was 15 of 31. Conversely in the Elephant and Castle 20 out of 25 said they felt safer.

Feelings about the area

Around half said the introduction of cameras had made them feel a little or much more positive about the area. A third felt it had made no difference. Again this varied with site. In East Street and Camberwell 41% believed it had made a positive difference compared with 64% in the Elephant and Castle. There was little variation between genders. Only 40% of 16-24 year olds felt more positive. Reasons given for feeling safer included: greater feelings of security; the activities of people are being watched over; the cameras gave confidence to shoppers; and less crime.

As with the surveys of the general public, about half of business respondents said they felt more positive about the area. Seven in ten surveyed in the Elephant and Castle maintained they felt more positive. This compared with 14 out of 31 in East Street.

Respondents were asked how they thought the introduction of CCTV had directly affected aspects of their business. About a third believed it had reduced levels of shoplifting and burglaries. Forty-seven percent (62) suggested they felt more secure at work as a result – and about a third said their feeling of security had increased going to and from work. Almost half (63) also felt that customer confidence had increased in the area.

Table 3.5 - Effect of CCTV on aspects of your business (n = 132)



About the same



Levels of shoplifting










Feelings of security at work





Feeling of security on your way to work





Customer confidence in the area





Amount of insurance paid





Volume of trade






Funding of schemes

Figure 3.2 indicates that generally – and not surprisingly - business respondents favoured local authority and central government funding CCTV systems in the future. Twelve said businesses should contribute – but all of these looked to do this on a partnership basis primarily with local authority involvement. Respondents in Camberwell were most in favour of local authority funding (32/41) and least in favour of businesses contributing to systems (1/41). Some traders went as far to say that all shops should have CCTV and that central government should make a contribution to this.

Figure 3.2 - Who should fund CCTV in the future? (n = 134)

bar2.gif (3687 bytes)

There were a large number of additional comments offered by respondents. The main themes flowing from these remarks included: the need for increased public awareness and promotion of CCTV; feedback from the council/police about the effectiveness of the schemes; more cameras being needed; and improved camera reliability and use of the cameras by operators. In East Street there were a number of negative comments about the system, such as:

‘Some member s of the public seem to think they are a waste of time in view of the amount of pickpocketing that goes on and some doubt the cameras are even working.’

‘Crime still occurs – such as pickpocketing’

The comments about pickpocketing fit in quite well with recorded crime figures, which show an increase in thefts from the person. Despite feeling a little safer one comment from a shopkeeper in Peckham was particularly disturbing:

‘I have had a business in Peckham for the past 11 years. The fear of crime is so intense that I cannot even spend one single day away without caring something is going to happen – especially when gangs of youths walk in … I have experienced hundreds of problems from shoplifting to violence and suffered mental trauma for the past 11 years.’

Other comments of interest include:

‘I think CCTV is a good idea, but also it would be a good idea to update local businesses on its effect. And to have a local police officer to keep in touch with the area.’

‘There should be more notices to remind the criminals that CCTV is in operation’

‘More publicity should be given about its existence’

‘I think there should be local consultation, as there was with the installation, to enable feedback to improve the service.’

‘I do think it’s a good idea, but am aware that it moves more crime to the fringes where the cameras are not present.’

‘The system has been in operation for over six months. No figures seem to have been published to convince all parties concerned of their overall success.’


This chapter examines the impact of CCTV on recorded crime in each of the four relevant sites. It starts with overview of crime and disorder in Southwark. It then offers an overall assessment of impact of CCTV on crime. Finally it provides a site-by-site analysis as follows:

  1. target area
    1. overall change in crime
    2. notable changes in crimes which we expected to be affected by CCTV
    3. interesting changes in other crimes
  1. buffer and comparison areas
    1. overall changes in crime
    2. notable changes in crime
  1. overview of findings in each site
  2. changes in disorder calls to the police (Camberwell and East Street only)

For both buffer zones and comparative sectors we have excluded crimes which took place within the target area. We have not done this for police divisions or the borough as a whole. Analysis of disorder calls to the police for the year prior to and following installation are found at the end of the sections on Camberwell and East Street. It should be noted that each system was installed at a different time. Therefore we have focused on varying ‘before’ and ‘after’ periods which span different years.[16]

Overview of crime and disorder in Southwark

Recorded crime in Southwark is decreasing. In the year ending March 1998 it was 9% lower than the previous year. Some of this fall may in part reflect changes in reporting and recording, but the underlying trend is almost certainly downwards.

Figure 4.1 - Breakdown of crime in Southwark April 97 – March 98 (n = 38,976)

Figure 4.2 - Breakdown of disorder calls to the police in Southwark April 97 – March 98 (n = 34,441)


The reduction in Southwark to the year ending March 1998 was slightly larger than the fall for the Metropolitan Police Force Area (down 8%) and considerably greater than for England and Wales (down 4%).

Disorder calls, about half of which are accounted for by disturbances, increased by 1%[17]. This marks a slowdown in disorder calls to the police which had been rising at a substantial rate in the early 1990s (Audit Commission, 1995). Seven out of ten of the

38, 976 crimes recorded in the twelve months ending March 1998 were against property. An additional 17% were violent offences and 7% involved drugs. Figure 4.2 provides more details.

Table 4.1 - Crime types selected for this study for the Borough of Southwark


Feb 96 – Jan 97

Feb 97 – Jan 98

Feb 98 – Jan 99

Change 98/99 over 96/97

% Change 98/99 over 96/97





- 584

- 9%

Vehicle Crime




- 1,839

- 18%

Street Crime




- 387

- 15%





- 511

- 10%

Other thefts





+ 19%

Criminal Damage




- 898

- 25%





- 3,323

- 10%


Over the entire period covered by the evaluation (Feb 96 – Jan 99), reductions in crime for the borough as a whole have again been significant (see Table 4.1). The largest reductions were in vehicle crime (down 1,839) and vandalism (down 898). Thefts in particular shoplifting increased markedly (up 896).

Impact of CCTV on crime – overall assessment

In our experience and that of other independent evaluators (Short and Ditton 1995; Brown 1995; Squires and Measor 1996) certain crimes are more likely to be deterred by CCTV than others. A previous study in Sutton (Bulos and Sarno 1995) found burglaries and criminal damage decreased by 47% and 42% respectively in the year following installation. Skinns (1996) similarly found in a study of Doncaster that burglaries and incidents of criminal damage fell by 25% and 32%, respectively. Conversely, previous studies have shown that other crimes, particularly violent offences, appear to be less affected by CCTV.Brown (1995:vi) writes

‘... the effect on personal crime is less clear. In large Metropolitan areas, the cameras have had very little effect on overall levels of assaults and wounding.’

Squires and Measor (1995:2) similarly conclude,

‘... the underlying figures, for the first six months of 1995 show little change (3.5% increase) and the incidents of recorded violence and disorder seem [as above] to be on the increase (up 24%).’

The results from the current study show that street crime, in particular robberies, which were cited as a major problem in all sites, decreased significantly. A high proportion of those interviewed in the public perception survey supported the view that CCTV was likely to impact upon street crime. Forty-seven percent (238) believed thefts from the person would be affected. Slightly less (45%) cited robberies as likely to be impacted upon.

Other crimes most likely to be affected include: burglaries, criminal damage, and vehicle crimes. When aggregated these crimes fell from 1,613 during the period Feb 96- Jan 97 to 1,208 in the year Feb 98 – Jan 99. This amounts to a fall of 25%.[18] Crimes of violence have decreased across all four sites. There was a 7% reduction between the periods Feb 96 – Jan 97 and Feb 98 - Jan 99. This reduction is predictably lower than the average fall across all crimes over the same period (down 10%). On a positive note ABH decreased by 31% (down 87) but serious assaults were in the main unchanged (46 against 47).

The reason why CCTV has less effect on violent offences has been largely unexamined. However, they are distinguished by characteristics of the specific nature of the circumstances, location, social context and mutually reinforcing expectations of group behaviour. Town centres have long been locations for social networking such as group outings to a public house or a teenage gathering in or around a fast food restaurant. A study of six police force areas indicated that one in five crimes of violence might be described as ‘street brawls’. One in eight were ‘pub brawls’ (Davidoff and Dowds, 1989). These types of social gathering often involve the consumption of alcohol. This can result in peer pressure, particularly amongst males, to engage in raucous behaviour, which escalates or provokes violent behaviour. Gatherings of the latter kind can be perceived as threatening to passers-by and can become the focus of behaviour which preys on ‘out of group’ members. In any case these types of gatherings are likely to be boisterous and driven by social expectation. Participants are therefore arguably less likely to respond to CCTV. In some cases behaviour may actually be directed towards attracting attention, and the arrival of the police might be part of the desired outcome. Far from being discouraged by the presence of CCTV, provocative and violent behaviour may, under some circumstances, actually be fuelled. In other cases the most likely impact of CCTV may be to shift these activities into side streets, or alternative non-CCTV locations.

Impact on recorded crime in the Elephant and Castle

This section examines changes in recorded crime in the year before and two years following installation (Feb 96 – Jan 99). The target area includes the whole of the shopping centre and those areas under camera surveillance in the surrounding locality. The area covered by recently installed cameras in Elephant Road have not been included as part of the target area. This is because they were installed in late 1998, almost at the end of the ‘after’ period

Recorded crime in the Elephant and Castle target area decreased in the ‘after’ period compared to Year -1 pre-CCTV. Figure 4.3 clearly shows the introduction of cameras coincided with sharp reductions in offences just before and following installation. It fell by 11% in the year following installation (491 against 436) – and was 17% lower in the following year (408). The most significant falls were in robberies (down by 67), thefts from the person (down by 28) and non-residential burglaries (down by 11).[19] Crimes we believed less likely to be affected, assaults and public order offences, fell from 66 to 41 in the year following installation – but increased to 77 in the subsequent year. This is in line with other independent evaluations, which suggest that CCTV has had little or no impact on crimes of violence over the longer term.

Figure 4.3 - Monthly recorded crime figures for the Elephant and Castle target area (Feb 96 – Jan 99)

Figure 4.4 - Changes in recorded crime in the Elephant and Castle target area (2/96 - 1/99)

In the year following installation recorded crime in the buffer zone fell (on Year -1 pre-CCTV levels) from 2,090 to 1,803. The following year it fell to 1,745, a decrease of 17% on pre-CCTV levels. Substantial decreases were recorded in thefts from vehicles (-111), criminal damage (–80), thefts of vehicles (-58) and robberies (-46). Changes in the comparison sectors were mixed. There was a fall in recorded offences in the Newington Sector (down 15%) but a small increase in the Borough Sector (up 3%). Recorded offences were down by 2% in the Southwark Division. For the Walworth Division and borough as a whole crime was down by 13% and 10% on Year -1 pre-CCTV levels. Table 4.2 summarises these findings.

Table 4.2 - Comparative changes in recorded crime – the Elephant and Castle


Year -1 pre-CCTV (2/96-1/97)

Year 1 post-CCTV (2/97–1/98)

Change Yr. 1 post on Yr. –1 pre

% Change

Year 2 post-CCTV (2/98-1/99)

Change Yr. 2 post on Yr. –1 pre

% Change

Target Area[20]








Buffer Zone








Comparison Sector









Comparison Sector









Comparison Division (all Southwark)








Comparison Division (all Walworth)








Borough as whole (all)








Recorded crime in the target area has fallen. However, levels of crime in the buffer zone and in the Newington Comparison Sector also experienced large reductions (15-17%). Given these notable falls it is difficult to unpick whether this is evidence of positive effect – with a diffusion of benefits from the target area to the buffer zone, or simply part of an overall downward trend within this area. The fall in robberies is especially encouraging, and suggests that CCTV had an impact on the crime which prompted the introduction of the system. At this stage it would be premature to claim conclusive results, however.

Impact on recorded crime in Peckham

As emphasised in the methodology section we have been unable to collect CRIS data for the year prior implementation. This means a full ‘before’ – ‘after’ analysis is not possible for Peckham. However, some police data have been made available. Allegations of street crime (robberies and thefts from the person) and burglaries committed within the view of the cameras have been collected for the before and after period. Crimes, which occurred out of the field of vision of cameras were excluded. The statistics collected for this exercise are unlikely to correspond directly with those covering the Peckham target area mapped out by the research team. [23] The Crime Management Unit in Peckham has also provided some ‘before’ and ‘after’ data. This is for the months of August and September in the two years prior and following camera installation. These are aggregate data for the town centre beat areas.

Figure 4.5 - Changes in Street Crime in Peckham (October 94 - September 97)

Street crime decreased sharply in the two years post-installation compared with the year prior to the introduction of cameras. In the twelve months following installation it fell by 61% (290 to 114 crimes) on the year prior to implementation.

Year 2 showed a much smaller fall of six crimes (114 to 108) on Year 1. Similarly there were falls in burglaries with entry points in view of the cameras. Between May and September 1995 there were 50 burglaries in view of the cameras. This fell to 30 in 1996 and 11 in 1997 for the same time periods.

Perhaps the best ‘before-after’ analysis we can provide is for the town centre beats. However, changes in the police computer software and operators mean the only reliable analysis we can provide is for the months of August and September, 1994-97. The period August/September 1994 and 1995 refer to the ‘before’ period. August/September 1996 and 1997 are the post-installation period.

key1.gif (1190 bytes)

Figure 4.6 Changes in recorded crime in the Peckham beat buffer zone (Aug/Sep 94 and Aug/Sep 95 against Aug/Sep 96 and Aug/Sep 97).

Recorded crime fell from 1904 in the ‘before’ period to 1583 for the ‘after’ period or 17%. The largest reduction was in street crime which fell from 397 to 130 offences. There were also reductions in vehicle crime, burglaries and other thefts. Although burglaries decreased by 5% there was a marked increase in non-residential burglaries (up by 71%). Residential burglaries fell by about a quarter. Similarly there were notable changes in different types of violent offences. Serious assaults decreased by half (down form 35 to 18) – but ABH increased by the same percentage (up from 81 to 122).

The only additional statistical analysis we can offer for Peckham is of trends in the target area, buffer zone and comparison area for the three-year period ‘after’ implementation. The data cover the period February 1996 (three months after the system went live) to January 1999.

Whether one might expect CCTV to have a progressive impact over time is arguable. On the one hand, the novelty – and initial publicity – of the scheme might result in an immediate but unsustained impact after implementation. On the other hand, if systems were used to good effect, one might expect offenders to respond progressively to the cameras’ presence. The evidence is somewhat supportive of the former viewpoint.

Recorded crime in the target area decreased from 1,103 for the period February 96 to January 97 to 1,001 in the following year, a fall of 9%. The largest decreases were in non-residential burglaries (down 59), theft from the person (down 26) and incidents of criminal damage (down 21). Street crime [24] decreased by 20 offences. This however, hides an increase of 6 street robberies. In the following year (98/99) recorded crime increased by 68 to 1069. Shoplifting (which we believed unlikely to be affected) increased for a second year rising by 43% to 337. Residential burglaries and other assaults also increased. Robberies dropped by 38 to 60 in total. Thefts of motor vehicles and ABH also had notable falls (see Appendix A for details).

Figure 4.7 - Recorded crime in the Peckham target area (Feb 96 – Jan 99)


In the two years following installation recorded crime in the buffer zone (beat areas1, 2 and 4 excluding the target area) fell from 3,564 to 3,469. Non-residential burglaries experienced the greatest decline (down 82) closelyfollowed by thefts of motor vehicles (down 75). There is some evidence to suggest that robberies have been geographically displaced to this area – increasing by 49 or 27%. In the following year there was a steeper decrease of 13% to 3,024. The largest reductions were in: criminal damage (down 138), ABH (down 102), criminal damage to motor vehicle (down 86) and robberies (down 42).


Table 4.3 - Comparative changes in recorded crime - Peckham


Year 1 post-CCTV

(2/96–1/ 97)

Year 2 post-CCTV

(2/ 97–1/ 98)

Change Yr. 1 post on Yr. 2 post

% Change

Year 3 post-CCTV

(2/98 – 1/99)

Change Yr. 3 post on Yr. 1 post

% Change

Target Area








Buffer Area








Comparison Sector (Peckham)








Comparison Division (all Peckham)








Borough as whole (all)








Levels of crime also fell in all the comparison areas. Over the three-year ‘after’ period decreases in the comparison areas have been greater than the target area.[25] This suggests it is unlikely that CCTV is having a continuing and progressive effect on crime in this area. However, we would need more data from the ‘before’ period to be sure.

Impact on recorded crime and disorder in Camberwell

This section will examine changes in recorded crime over a three-year period in Camberwell (two years prior and one year following installation). The specific period under examination runs from February 1996 to January 1999. The target area in Camberwell primarily covers the town centre, particularly Camberwell Church Street and Denmark Hill (see Appendix D for full details).

Recorded crime has fallen in the Camberwell CCTV target area – both before and after the installation of CCTV. In the year before the system went live, there were falls in vehicle crime (-46), sex and violence offences (-30) and street crime (-25). However, notable increases in shoplifting (+48) meant the overall crime figure was only 4% lower in post-CCTV year 1 than the previous year (Feb 96 – Jan 97).

Figure 4.8 - Recorded Crime in the Camberwell target area (Feb 96-Jan 99)



Following installation a steeper drop of 12% was recorded (Yr. 1 compared with Yr. -2). Reductions in vehicle crime (down by 28) and street crime (down by 34) continued, albeit at a lower level in number terms than the previous year (down 101 compared with down 79). Burglaries, in particular those against non-residential premises (down by 11) also fell by 19. In examining specific crimes the single largest fall was in robberies, which decreased by 22. Surprisingly sex and violence offences decreased (down by 17) for a second year running. There were increases in bicycle and other thefts (up by 12) as well as serious assaults (up by 3). A fuller breakdown of year-on-year changes for specific crimes is set out in Appendix A.


Figure 4.9 - Changes in recorded crime in the Camberwell target area (Feb 96 – Jan



Table 4.4 - Comparative changes in recorded crime -Camberwell


Year -2

pre-CCTV (2/96-1/97)

Year -1



Change Yr. 1 pre on Yr. 2 pre

% Change

Year 1 post-CCTV


Change Yr. 1 post on Yr. –1 pre

% Change

Target Area








Buffer Zone








Comparison Sector (Camberwell)








Comparison Division (all Walworth)








Borough as whole (all)









During the pre-installation period crime in the buffer zone decreased from 1,354 to 1,136 [26]. However, this fall was followed by an increase from 1,136 to 1,210 (or up 7%) in the year post-installation. There were increases in cases of criminal damage to motor vehicles (+25), ABH (+25) and section 4 public order offences (+9). The most worrying increase, however, was in burglaries, which went up by 39% from 228 to 316. Residential burglaries rose from 187 to 262 (29%). Nevertheless, there were also notable falls in street crime (-48)[27], criminal damage (-19) and serious assaults (-5) - (See Appendix B for more detailed figures).

Recorded crime in the Camberwell Sector rose by 3% (3653 to 3755). There were substantial rises in burglaries (+79 or +10%), thefts of motor vehicles (+70 or +24%) and ABH (+52 or +25%). Figures for the Walworth Division and borough decreased by 6% and 5%, respectively.

Clearly recorded crime in the target area decreased quite significantly in the year following CCTV installation (-12%). Promising falls in street crime, vehicle crime and violent crime were recorded although these had already been decreasing significantly prior to the introduction of cameras. However, it is probable that these falls have been at the expense of the surrounding buffer areas where crime has increased by up to 7%. Most disconcerting of all is that crimes, which were falling in the target area have increased significantly in the beat and sector buffer zones. Burglaries and violent crime in the buffer zone, for example, increased by 39% and 26%, respectively.

Disorder Calls to the police in Camberwell

In Camberwell disorder calls to the police increased from 815 to 846, or 4%. The most significant increase was in abandoned phone calls to the police, which increased by 28% from 208 to 266. Disturbances in public places (the largest single category) increased slightly from 414 to 419. There were notable falls in disturbances in licensed premises (down by 15) and civil disputes (down by 16).

In the year following installation disorder calls in the buffer zone increased at three times the rate of the target area. In number terms there was an increase of 162 calls, or 14% (1331). As in the target area the biggest rise was in abandoned phone calls which increased from 353 to 507. The biggest decrease was in noise nuisance calls, which fell from 74 to 29. Civil disputes and domestic incidents fell by 8 and 9 calls respectively. Overall, disorder calls increased by 4% in the Camberwell Sector – but were unchanged in the Division as a whole. In Southwark there was a 1% increase in calls.

Impact on recorded crime and disorder in East Street

Cameras were introduced into East Street in January 1998. We have therefore examined the three-year period February 1996 to January 1999 – or two years prior to installation and one year following. The system covers a large tract of East Street running from the Walworth Road to Elsted Street, thus covering the majority of the outdoor market area. As with other CCTV schemes introduced into Southwark, cameras have been placed at junctions allowing some coverage of adjacent roads. Particularly important is the camera at the top end of East Street, which allows surveillance of the Walworth Road from Browning to Merrow Road.

Figure 4.10 - Recorded crime in the East Street target area (Feb 96 – Jan 99)

Recorded crime in the target area increased by 3% (up from 780 to 801) in the year prior to CCTV installation (Yr. -2 pre-CCTV compared with Yr. -1 pre-CCTV). Figure 4.10 (below) indicates that the majority of this rise was in vehicle crime (up 23). In the subsequent year, however, the overall figure fell to 717, a reduction of 10% on the previous year. There were notable decreases in vehicle crime (down 44) and criminal damage (down by 23). Street crime rose sharply from 156 to 190 – but this overall figure conceals significant variations by crime type. Robberies almost halved (down from 50 to 26) but theft from the person increased from 106 to 164. This shift might represent functional displacement from one crime type to another. Thefts from the person may have increased as it is a subtler and less obvious criminal act and therefore is more likely to go unnoticed by CCTV camera operators. Violent and sexual crimes were largely unchanged (111 to 109) – but worryingly there was an increase in more serious offences including ABH (up from 36 to 43) and serious assaults (up from 11 to 16). Other thefts, which include shoplifting, unexpectedly fell by 42 offences. (See Appendix A for more detailed analysis of changes).

Figure 4.11 - Changes in recorded crime in the East Street target area (Feb 96 – Jan 99)


Table 4.5 - Comparative changes in recorded crime - East Street


Year -2

Pre-CCTV (2/96- 1/97)

Year -1 pre-CCTV (2/97–1/98)

Change Yr. -1 pre on Yr. -2 pre

% Change

Year 1

Post-CCTV 2/98-1/99

Change Yr. 1 post on Yr. –1 pre

% Change

Target Area








Buffer Zone








Comparison Sector









Comparison Division (all Walworth)








Borough as whole (all)








In the year preceding installation there was little change in recorded crime in the buffer zone (beats 3 and 5 in the Newington Sector excluding the target area). Following the introduction of cameras, however, recorded offences fell from 1,058 to 830, or 22%. This reduction is significantly greater than the target area. Large falls were recorded in thefts from motor vehicles (down 77), shoplifting (down 42), common assaults and assaults on police officers (down 37) and criminal damage (down 35). One notable exception was a significant rise in thefts of motor vehicles, which increased by 32 to 104. Levels of crime in the Comparison Sector (Newington) fell by 12% or 488 offences. This is mainly as a result of sharp decreases in thefts from motor vehicles (down 184), robberies (down 120), shoplifting (down 81) and theft of motor vehicles (down 81). Figures for the Walworth Division and borough fell by between 56% and 5%.

Following the introduction of CCTV, recorded crime in the target area decreased by 10%. This is however much lower than the buffer zone where there was a 22% decrease. We cannot fully explain this pattern. We can say however that we do not believe reductions in recorded crime within the buffer zone and comparison sector are the direct result of a diffusion of benefits from the East Street system. Nevertheless, vehicle crime and other thefts – particularly shoplifting have fallen markedly. The increase in serious violent offences is worrying and we have no immediate explanations for this. We have no reason to believe there has been any geographical displacement from the target area to the buffer zone. However, it is possible that functional displacement within the target area has occurred – thus accounting for the decrease in robberies and increase in thefts from the person.

Disorder calls to the police in the East Street area

There was a fall in disorder calls in the East Street target area. The overall decrease was 9% or 90 calls. There were decreases in all disorder categories. In number terms the largest falls were in disturbances in public places (down 18), disturbance in private premises (down 17) and domestic incidents (down 15). In the buffer zone the decrease equated to 10 incidents, or about 1%. The steepest decreases were in noise nuisance and domestic incident calls (each down by 19). Figures for the Newington Sector and Walworth Divison were largely unchanged. Disorder calls for the borough as a whole increased 1%.

Cost effectiveness

As has become clear, it is complicated to unravel precisely what impact the four CCTV schemes have had on recorded crime. This is partly because we lack full data on one site; but even where we have adequate data, trends in target areas, buffer zones and comparison sectors do not tell a clear story. It seems very likely that the schemes have impacted on street robbery; it is less clear that they have done so on other categories of street crime, and other forms of crime. Whatever the impact on recorded crime, we must be even more circumspect about the effects on unreported and unrecorded crime. Some of the crimes likely to be affected by CCTV go largely unreported to the police.

Whatever the complexities in establishing impact, assessing the systems’ cost effectiveness is relatively straightforward, provided that one starts with a reliable estimate of crimes prevented. One can calculate annual costs by charging a notional annual interest on capital costs and summing this figure with running costs. The cost per crime prevented is calculated simply by dividing the annual cost estimate by the number of crimes prevented.

Capital costs for the four systems were:

  • Camberwell£307,000
  • East Street£168,000
  • Peckham£366,000
  • Elephant and Castle£310,000

Summing these and setting an annual cost at 10% of the total yields a figure of £115,000 per year. The annual running costs for the two control rooms is around £165,000. Thus the total annual costs are in the region of £280,000.

How cost effective is this expenditure? It depends entirely on the number of crimes prevented, and on other related benefits. If one assumed that 70 recorded crime were prevented annually per site, the cost per recorded crime prevented would be £1,000. It is also important to make some assumptions about unrecorded crime. If 50% of crimes prevented reach police statistics, the cost per crime prevented would fall to £500. It is equally important to take account of the benefit of reduced fear, though it is hard to see how this could be included in any cost effectiveness analysis.

In assessing whether this expenditure is good value for money, one needs to assess the cost of preventing crime by other means. Cost effectiveness analysis in criminal justice is in its infancy, and we know of no relevant work. One can obviously offer some indicative figures. For example, if one attributed have the police budget in Southwark to crime reduction in its broadest sense, there is roughly £2,000 spent per recorded crime. What we don’t know is the marginal costs per crime prevented. Again, one can estimate the costs of reducing crime through incapacitation by imprisonment: if seven recorded crimes are prevented for each complete year of imprisonment, the cost per crime prevented is around £4,000.

We have presented these figures simply to illustrate the likely costs per crime prevented through CCTV and alternative strategies. We cannot pretend that our estimate of crimes prevented is anything more than an informed guess. However, the exercise suggests that investing in CCTV could easily make sense, especially when other benefits besides crime prevention are included in the calculus.



Summary of findings

Across all sites crime deterrence (in particular street crime) and a reduction in the fear of crime featured consistently in the primary aims of each system. In the case of the Elephant and Castle, Camberwell and East Street there were reductions in some recorded offences[28], particularly robbery. Street crime in both the Elephant and Castle and Camberwell showed falls of 23% in the year following installation. East Street, however, recorded a 22% increase in overall street crime. In all sites between 53% and 69% of the public believed they felt safer now that the cameras have been installed. Nevertheless, in spite of increased feelings of safety across all sites less than half of respondents maintained they felt safe after dark.

Overall results from this evaluation point to reductions in recorded crime in the order of 10% to 12% in the year following the introduction of cameras[29]. Whilst these reductions are impressive they must be taken in the context of falling levels of recorded crime across the borough. In the year ending January 1999 recorded offences in Southwark were 5% lower than the previous twelve months. Also, with the exception of Camberwell, reductions in the buffer zones (police beats surrounding the target areas) have been equal to or greater than those areas under direct camera surveillance. Changes in recorded crime in the comparison areas were mixed. There were increases in the number of offences in the Borough and Camberwell Sectors (up 2% and 3% respectively) in years following installation. The Southwark and Walworth Divisions however recorded falls of between 2% and 8%. Trying to unpick and make sense of these figures is difficult. Although falls in the target areas have outstripped those of comparison areas – buffer zones have generally performed equally as well as target areas. Our view is that with the exception of Camberwell it is difficult to directly attribute reductions in crime within the target areas unequivocally to the introduction of CCTV. This is primarily because of the equally substantial reductions across buffer zones. It is more likely to be part of a general downward trend in crime within which CCTV has played a minor role.

Detections through the use of CCTV have been relatively low except for the Elephant and Castle. Between June and September 1998 only eight tapes were taken for examination by police from the Cerise Road control room (which monitors the Camberwell, Peckham and East Street systems). Staff at the Cerise Road control room scan Peckham, Camberwell and East Street with the cameras - but in what might be described as a blind fashion. They receive little or no intelligence information on who to look out for and have no formal training on what to look for – i.e. what constitutes suspicious behaviour. As a direct consequence very few incidents ever come to the attention of controllers. When an incident is caught on camera, we believe it is usually an unmanned camera and thus not zoomed in to capture good clear evidence.

In contrast, police removed 68 tapes from the control room in the Elephant and Castle – although about half of these were taken during or just after the Stephen Lawrence enquiry. Nevertheless even when tapes seized for the enquiry are excluded an appreciably greater number of tapes were removed from the Elephant control room (37 against 8). We believe the relative success of the Elephant and Castle Scheme in deterring and detecting crime can be attributed to the introduction or presence of other measures. These include: regular information exchanges between home beat officers and control room staff; ‘Operation Watch’ picture file established; security patrols in the shopping centre; and local businesses linked up to the control room. Despite the success of the Elephant scheme in terms of crime reduction and detection there is a need for the scheme to be systematically reviewed in the light of the codes of practice and civil liberties issues. (See good practice lessons and control room management recommendations)

Our findings support the need for the introduction of closed circuit television as part of a package of measures. Success is simply not a matter of installing reliable hardware. Of particular importance is the need for staff training in what to look for and sources of information to alert controllers of suspicious behaviour or particular incidents. The introduction of ‘town centre wardens’ may help to fill the dearth/void of information flowing to CCTV controllers. This coupled with regular information sharing with the police is likely to not only make the job of controllers more interesting but also more productive in terms of detections. There is also scope for changing shift and break patterns. Although we understand that shifts patterns at the control room are standard across the security industry we feel they are too long and breaks too short.

Good Practice Lessons

Cerise Road control room

  1. Formal police training for controllers on what to look for and not to look for
  2. Greater levels of co-operation between control room staff, police and shop security by establishing a Business Watch scheme in Peckham and Camberwell
  3. More frequent visits by home beat officers to the Cerise Road control room to share information (as happens at the Elephant)
  4. An Operation Watch picture file established at Cerise Road provided other recommendations on control room practices are implemented
  5. Improved two-way communication between control rooms and CAD rooms. A hotline established between Cerise Road and the CAD room at Peckham
  6. Continued use of police covert cameras in conjunction with main systems where applicable

Elephant and Castle control room

  1. The Elephant and Castle scheme straddles boundaries of two police divisions. There is scope for greater levels of co-operation between relevant home beat officers in the two divisions. We believe that home beat officers from both beats could benefit from greater levels of contact with one another and shared knowledge
  2. Increased use of the Elephant and Castle system by the Southwark Division – more visits by relevant beat officers and the sharing of information with controllers

All sites

  1. The presence of all systems to be more highly publicised. This could be done through improved signage, more shop window posters, and successful detections highlighted in local newspapers
  2. The police given information on the role of CCTV controllers

Continuing Evaluation

Time periods examined in this evaluation have been short. For example we have only assessed changes in recorded crime over a one-year ‘after’ period in Camberwell and East Street. To examine whether CCTV is having a progressive effect data needs to be continually collected and procedures put in place for longer term monitoring. We would recommend that:

  1. The police keep formal records of detections and convictions resulting from use of cameras - successes should be publicised where possible in the local press
  2. Questions about CCTV should be inserted into the MORI poll of Southwark residents
  3. Continued annual monitoring of recorded crime in the target, buffer and comparison areas
  4. Continual review of the code of practice and monitoring of its implementation
  5. Dissemination of results to local residents and businesses

Control Room Management

In the light of the possible implementation of picture files at Cerise Road we believe there is a need for tighter control room management at both sites. This is to ensure strict adherence to the Southwark code of practice.

  1. Limited access to controllers to view tapes on their own
  2. Strict controls on who is allowed to produce stills and under what circumstances
  3. A system of recording and accounting for still photographs
  4. Controls to ensure video tapes are signed out by police officers. Logbooks should be checked against tapes being stored to ensure this has occurred
  5. Regular management spot checks
  6. Some repairs seem to be taking longer than one would expect. Camera 4 in Peckham to be sorted out as soon as possible. Borough technicians/or outside body to review system regularly
  7. A review of the code of practice
  8. Training in respect of code of practice ‘good practice’


Brown, B. (1995) CCTV in Town Centres: Three Case Studies, Home Office Police Research Group, Crime Detection and Prevention Series, Paper 68. HMSO

Bulos, M and Sarno, C (1994) Closed Circuit Television and Local Authority Initiatives: The First National Survey, School of Land Management and Urban Policy, Research Monograph. RM1/1994, South Bank University, London.

Bulos, M ed. (1995) Towards a Safer Sutton? Impact of Closed Circuit Television on Sutton Town Centre, London Borough of Sutton.

Bulos, M. and Sarno, C. (1996) Codes of Practice and Public Closed Circuit Television Systems, Local Government Information Unit. London.

Davidoff, L. and Dowds, L (1989) Recent trends in violence against the person in England and Wales, Home Office Research and Planning Unit, Research Bulletin, No. 27, HMSO

Ditton, J. et al (1999) The effect of closed circuit television on recorded crime rates and public concern about crime in Glasgow, Scottish Office Central Research Unit: Edinburgh

Local Government Information Unit (1996) A Watching Brief, Local Government Information Unit. LGIU: London.

Short, E. and Ditton, J (1995) Does CCTV Affect Crime, CCTV Today 2/2: 10-20.

Short, E. and Ditton, J (1995) Does closed circuit television prevent crime? An evaluation of the use of CCTV surveillance cameras in Airdrie town centre, Central Research Unit, The Scottish Office: Edinburgh: HMSO

Skinns, D (1997) Annual report of the Safety in Doncaster evaluation project, October 1995-September 1995, Doncaster Council & South Yorkshire Police in Partnership.

Squires, P and Measor, L (1996) Closed Circuit Television Surveillance and Crime Prevention in Brighton Half-Yearly Report. Brighton Health and Social Policy Research Centre, University of Brighton: Brighton: U of B.

Tilley, N (1993) Understanding Car Parks, Crime and CCTV: Evaluation Lessons from Safer Cities, Home Office Police Research Group, Crime Prevention Unit Series, Paper 42, London: HMSO.


  1. To maintain anonymity Omnidata substitutes postcode centroid grid references for the address grid reference.
  2. Data for police sectors excludes crimes which occurred in the target area
  3. Table 1.1 does not include the borough as a whole, though we also present borough-wide trends.
  4. Recorded crimes based on allegations of offences.
  5. Note of caution – It is likely that in the changeover period to CRIS (February and March 1996) it is probable that there was some under-recording of crime figures
  6. CRIS codes in brackets
  7. Including the pilot survey 851 interviews were completed. 227 in Peckham; 224 in the Elephant
  8. "Before-and-after" crime surveys could be conducted, but samples would have to be very large indeed – and thus very costly – to be sensitive to small falls in rare crimes.
  9. For example, the 1996 British Crime Survey Statistical Bulletin (Mirrlees-Black et al., 1998) estimated that there were seven times as many offences of vandalism and eight times as many robberies and thefts from persons as were actually recorded by the police.
  10. Viewing of tapes from cameras outside the shopping centre contravenes the code of practice.
  11. Southwark Council contributed to 90% of the capital and 75% of running costs of the scheme.
  12. The police do ask operators to focus cameras on certain addresses or areas for operations but do not give details.
  13. Controllers noted that radio links are being used more frequently. They however felt shop security staff needed more training in using the radios to relay information.
  14. It should be noted that the Cerise Road control room monitors three systems and is therefore more likely to have a larger number of faults reported. Furthermore, systems in Camberwell and East Street had been operational for less than 6 months and were thus were still likely to be suffering from ‘teething difficulties’.
  15. 56% lived outside of Peckham and 38% used the area once a week or less.
  16. The reader should refer to Table 1.1 in the Methods section to be clear about these issues.
  17. Public order instances – i.e. disturbance in public place or disturbance in licensed premises
  18. Aggregated across the 4 target areas
  19. Feb 96 – Jan 97 compared with Feb 98 – Jan 99
  20. The target area in the Elephant and Castle straddles the boundaries of the Walworth and Southwark Divisions.
  21. Excludes all recorded offences which occurred in the shopping centre. It also includes half of crimes which happened on-street in the target area.
  22. Excludes half of crimes which occurred on-street in the target area.
  23. Allegation statistics were examined on a case-by-case basis, and a decision made as to whether they fell within the area under camera surveillance or not.
  24. Street crime comprises thefts from the person and robberies in open public space.
  25. One must remember that we are not comparing the ‘before’ and ‘after’ installation periods. These figures relate specifically to the after period only.
  26. Year 1 pre-CCTV compared with Year 2 pre-CCTV
  27. Robbery down from 113 to 78 and theft from the person reduced from 34 to 21.
  28. We cannot say anything about changes in recorded crime in Peckham as we have incomplete pre-installation data.
  29. The Elephant and Castle scheme showed a progressive affect. Recorded crime fell from 491 offences in the year prior to installation to 436 in Yr. 1 post installation to 408 in Yr. 2 post- installation.


FHSS home |  Courses | Research | Divisions |  Events |  Find |  SBU Home

This page was last updated by FHSS webmaster
on 28 April 2000