Promising - One study
Date: This profile was posted on June 08, 2011
This program installed closed-circuit television cameras to monitor public space in different locations around the city of Philadelphia, PA. This program is rated Promising. The intervention showed a statistically significant reduction in disorder crime and the number of crime events in the target areas, compared with the control areas. However, no significant differences were found for serious crimes in the target areas, compared with the control areas.
The city of Philadelphia, Pa., installed 18 closed-circuit television (CCTV) cameras at various locations in the city between July 2006 and November 2006 to help reduce crime. These cameras are monitored by the Philadelphia Police Department.
Of the 18 cameras, eight were pan, tile, zoom cameras. These cameras can be controlled by a remote operator so that they can tilt up and down, pan around the surrounding area, or zoom in on a particular segment of the viewing area or “viewshed.” These images are viewed in real time by a police officer. Images are recorded digitally and stored for up to 12 days. These cameras provide good-quality images that allow a license plate to be read up to a block from the camera or for street activity to be viewed up to three blocks away. In Philadelphia, the officer monitoring these images is located at police headquarters.
The remaining 10 cameras were Portable Overt Digital Surveillance System cameras. They do not allow for real-time image viewing. These cameras are moveable, the recording system is located in a bullet-resistant unit, and the camera emits a flashing strobe light to attract the attention of people in the area. The images can be viewed by patrol officers with equipment designed to access a wireless link. Images can be stored digitally for up to 5 days. The retrieval of images from these cameras can be time-intensive: it requires the help of a city street engineer and can take up to two hours.
The underlying theory for the effectiveness of CCTV is deterrence theory, which predicts reductions in crime resulting from potential offenders’ perceptions of increased risk of detection and capture (Ratcliffe, Taniguchi, and Taylor 2009).
Ratcliffe, Taniguchi, and Taylor (2009) found that camera implementation had no significant impact on the amount of serious crime in the target areas, compared with the control areas.
Camera implementation resulted in a statistically significant reduction in disorder crime in the target areas. After camera implementation, the average expected disorder crime count for the target areas was 16 percent lower than the control areas.
All Crime (Serious and Disorder Combined)
After implementation of cameras, there was a statistically significant reduction in the number of crime events within the target areas. In the months following the implementation of the cameras, there was a 13.3 percent reduction in expected crime counts in the target areas, compared with the control areas.
Ratcliffe, Taniguchi, and Taylor (2009) used a quasi-experimental design that incorporated “before” and “after” measures of crime in experimental and comparable control areas. They used reported crime data independently collected by the police. These crimes were divided into two categories, “serious” and “disorder” crimes; crimes that could not be affected by CCTV were excluded from the analysis. The authors noted the low volume of serious crime at each camera site, which made it difficult to detect significance for the study's findings. The follow-up period was nine months for most sites (for a few sites, the period was slightly longer).
Although the 18 cameras were installed at 12 sites, the close proximity of some of the cameras led the researchers to construct eight evaluation sites. The treatment areas consisted of those areas that mapped the actual viewsheds of the cameras. Buffer areas surrounded these viewshed areas (approximately a 500-foot zone), and were the places to which crime would likely be displaced or, alternatively, that would benefit from a diffusion effect. The control areas were defined as the police districts surrounding the displacement areas.
The study assessed a time series of monthly crime incidents for 32 months for eight locations. The data was analyzed using hierarchical linear modeling and weighted displacement quotient. Controls were incorporated for seasonality and long-term trends in the crime data. Only crimes occurring in the carefully defined viewshed of cameras were included in the analysis for main effects. The analysis accounted for the displacement/diffusion of benefits.
There is no cost information available for this program.
Evidence-Base (Studies Reviewed)