Eric L. Piza, Brandon C. Welsh, David P. Farrington, & Amanda L. Thomas (2019)
Criminology & Public Policy, 18(1): 135-159
*This study was funded by the Swedish National Council for Crime Prevention
- CCTV was associated with a ~13% reduction of crime in target areas as compared to control areas
- Effects were strongest in car parks followed by residential areas
- Of the countries where CCTV evaluations were conducted, the strongest effects were observed in United Kingdom and South Korea
- Schemes incorporating active camera monitoring had larger effects than passive systems
- Schemes deploying multiple interventions alongside CCTV had larger effects than schemes deploying single or no other interventions
This study updates prior systematic reviews and meta-analyses testing the effect of CCTV video surveillance cameras on crime. Systematic reviews incorporate rigorous and transparent methods for locating, appraising, and synthesizing scientific evidence from prior evaluation studies. Meta-analysis is a statistical technique that combines results from multiple evaluations into a single weighted average effect size.
We searched for CCTV evaluations published from 2007 through 2017 to account for the time period since the last review. The authors identified 68 CCTV evaluations from this time period. Considering these evaluations alongside those identified in the prior CCTV reviews, this study included a total of 161 CCTV evaluations conducted between 1978 and 2017. 80 evaluations met the inclusion criteria with 76 providing the necessary data for the meta-analysis.
Overall, CCTV was found to have a modest but significant effect on crime. Crime decreased by ~13% in CCTV areas as compared to control areas.
CCTV effect was compared across six geographic settings: car park (n=8), city/town center (n=33), housing complex (n=10), residential area (n=16), public transport (n=4), and other setting (n=5). The largest effects were observed for car parks, with CCTV areas experiencing a significant crime reduction of ~37% as compared to control areas. A significant reduction of ~12% was observed in residential areas as compared to control areas.
The analysis also tested whether the effect of CCTV differed across countries: United Kingdom (n=34), United States (n=24), Canada (n=6), South Korea (n=3), Sweden (n=3), Norway (n=1), Spain (n=1), Poland (n=2), and Australia (n=1). For the meta-analysis, Norway, Spain, Poland, and Australia were grouped together as “other country.” Significant reductions of approximately 21% and 34% were observed for the United Kingdom and South Korea, respectively. However, the United Kingdom findings did not lose significance when a more conservative effect size was used while the South Korea findings lost significance with the more conservative measure. Additionally, the South Korea findings warrant caution given the small number of studies conducted in this country (3) as compared to the United Kingdom (34). No other country exhibited any significant effects.
65 studies reported the type of monitoring used in CCTV schemes as active (n=54) or passive (n=11). Active systems were associated with an approximate 15% reduction in crime while no significant effects were observed for passive systems. Schemes reporting the deployment of multiple interventions (n=14) alongside CCTV generated an approximately 34% reduction of crime in treatment areas as compared with in control areas. No significant effects were observed for schemes reporting either none (n=36) or one other intervention (n=26) alongside CCTV.
The increase in CCTV evaluations has improved the knowledge base around the crime prevention effect of the technology. While overall findings suggest a modest but significant effect on crime, important geographic and strategic aspects of CCTV schemes influence the likelihood of success. Public safety agencies should take these considerations into account when designing their CCTV programs. Future research should aim to incorporate rigorous evaluation designs to ensure the policy relevance of CCTV evaluation studies.