• Center for Problem oriented policing

POP Center Responses Video Surveillance of Public Places, 2nd Ed. Page 3

Evaluations of CCTV              previous pagenext page

Several surveys have examined the perception of CCTV system managers and the public in regard to CCTV’s crime prevention benefits.35 These perceptions are usually positive, but evidence of actual crime reduction is harder to come by. In the early days of CCTV, many evaluations were carried out, but a number of significant methodological considerations draw into question their reliability. Problems included a lack of control areas, independence of researchers, and simplistic approaches to temporal crime patterns. 

Establishing if CCTV reduces crime is often difficult because a problem-oriented policing solution is rarely implemented without incident or without other crime prevention measures being applied at the same time. The implementation can often run into problems and commence late or in piecemeal fashion; crime rates naturally vary and show evidence of seasonality and long- and short-term trends; offenders are not necessarily aware of the system or become aware at different times (a theoretically crucial mechanism to CCTV success); and there are quantitative challenges to the measurement and detection of displacement and diffusion of benefits. These issues make it difficult to detect the impact of CCTV alone. For example, although CCTV was a factor affecting the operation of four street drug markets in London (UK), the cameras were often used with other crime prevention/detection efforts, such as large-scale arrests of sellers and situational-crime-prevention measures.36 

In some cases, the sheer lack of crime inhibits any robust evaluation. For example, the state of Illinois is reported to have spent $4 million installing cameras at all interstate rest areas. The cameras are monitored by state police. However, both the Illinois Department of Transportation and the state police admitted that serious crime at rest areas is extremely rare, with the latter identifying about 50 total crimes per year at all rest areas in the state.37 With such low crime rates, it may be impossible to demonstrate any crime-reduction benefit for the millions spent. 

Assessing the impact of CCTV is also complicated by the system’s design. CCTV is designed to see crime. As a result, the cameras may detect offenses that police would not otherwise notice. This may inadvertently increase the reported-crime rate, especially for offenses that have low reporting rates. In the United States, the reporting rate of violent crime is only 41 percent.† A process by which police can become aware of street violence without having to rely on the cooperation of the general public may increase reporting rates substantially. This does not mean crime will go up, but it is possible recorded crime may rise, as was probably the cause for a significant increase in reported woundings and assault in more than one UK town.38

† Only 40.9% of respondents said they had reported a violent victimization to the police (Bureau of Justice Statistics, 2020, Criminal Victimization, 2019, Table 6 at www.bjs.gov/content/pub/pdf/cv19.pdf). 

There have been a number of evaluation reviews (Appendix B) and more recently a number of meta-analyses (Appendix A). Phillips39 concluded that CCTV can be effective against property crime, but the results were less clear regarding personal crime and public order offenses, and the results were mixed in regard to reducing fear of crime. Similarly, Welsh and Farrington’s meta-analysis of 13 programs found five that appeared to work, three that appeared not to, and five that produced inconclusive results.40 Piza and colleagues found in their meta-analysis overall modest significant reductions in crime with limited displacement and some diffusion of benefits, yet the results differ based on location and type of monitoring. Gill and Spriggs41 evaluated 13 UK CCTV systems, finding that six demonstrated a relatively substantial reduction in crime in the surveilled area when compared to the designated control area. Of these six, only two showed a statistically significant reduction relative to the control zone.‡ In seven areas there was an increase in crime, though the increase could not be attributed to CCTV. Other potential causes for the crime increase included fluctuations in crime rates caused by seasonal, divisional, and national trends, and additional initiatives.

‡ And as the report authors note, “in one of these cases the change could be explained by the presence of confounding variables.” 

The evaluations in Appendix B and meta-analyses in Appendix A go some way to confirming these rather confusing findings. The general findings suggest that: 

  • CCTV is more effective at combating property offenses than violence or public order crime (though there have been successes in this area)
  • CCTV appears to work best in small, well-defined areas (such as public car parks)
  • The individual context of each area and the way the system is used appear to be important
  • Active systems are better at reducing crime than passive systems
  • Within one CCTV system there may be differences in effectiveness depending on the location
  • Achieving statistically significant reductions in crime can be difficult (i.e., crime reductions that clearly go beyond the level that might occur due to the normal fluctuations in the crime rate are difficult to prove)
  • CCTV systems may be able to diffuse benefits to buffer areas surrounding the cameras
  • A close relationship with the police appears important in determining a successful system
  • There is an investigative benefit to CCTV once an offense has been committed 

Reading this, you may feel the answer is unclear. Academic evaluators tend toward caution in their language, as they understand there is often a complex pattern of factors that dictate whether a system is successful or not. The rigid requirements of statistical evidence often limit the conclusions that quantitative evaluators can draw. 

To move beyond a strictly statistical interpretation, it is possible to say there was some evidence of crime reduction in most of the systems reported in the appendices. In other words, CCTV will almost certainly not make things worse (though crime reporting may increase), and there is a growing list of evaluations that suggest CCTV has had some qualified successes in reducing crime. 

The important point is that the local context is central to determining the likelihood of success. For example, city streets with long, clear lines of sight may be more amenable to CCTV than short, narrow winding lanes with trees that might obscure camera views. The availability of police to respond to incidents in an appropriate manner may also be a local context that affects CCTV’s success. Areas with high levels of property crime may be more amenable to CCTV than areas with low levels of public disorder. Smaller systems in well-defined areas may be more effective than broad-ranging systems that cover large areas. Understanding your local context is central to a successful problem-oriented policing solution. 

CCTV appears to be somewhat effective at reducing fear of crime, but only among a subset of the population. There are examples of a reduction in fear of crime among some people who are in CCTV areas, but it requires them to know they are in a surveillance area, and this is often not the case. Relying on CCTV to reduce fear of crime may require a significant and ongoing publicity campaign.§ Reduction in fear of crime may also differ based on gender so campaigns may also need to focus on women who are less likely to have their fear reduced by CCTV.

§ See Problem-Solving Tools Guide No. 5, Crime Prevention Publicity Campaigns for further information.

previous pagenext page