Is the Punishment More Certain? An Analysis of CCTV Detections and Enforcement

by Piza, E., Caplan, J. and Kennedy, L. (2014)

Justice Quarterly, 31(6): 1015-1043

Research Brief
Open Access Post Print
Published Article

*This study was funded by the National Institute of Justice (grant number 2010-IJ-CX-0026)

Key Takeaways

  • Overall, calls for service exhibited shorter dispatch processing times than CCTV detections
  • CCTV detections experienced higher case closure rates than calls for service for 6 of 8 crime types
  • Surveillance activity steadily (and drastically) decreased over the study period
  • Each new phase of camera installation decreased weekly levels of both CCTV detections and enforcement by 47%
  • Police officer layoffs decreased weekly detections by 86%
  • The introduction of an acoustic gunshot detection system was associated with a 29% reduction in weekly CCTV detections
  • Overall, this study supports the notion that CCTV increases certainty of punishment
  • Increased effectiveness is largely negated by CCTV detections and enforcement becoming rare as the system expanded

Research Summary

The primary preventive mechanism of CCTV video surveillance is considered to be deterrence. It is often taken for granted that the simple presence of cameras increases the certainty of punishment—a key ingredient of deterrence—for crimes observed on CCTV. However, empirical findings from the research literature cast doubt on this assumption. Numerous studies have documented offender willingness to commit crime in sight of CCTV, largely owing to the perception that the presence of a camera does not guarantee criminal infractions will result in enforcement.

“Surveillance barriers” inherent in the practical deployment of CCTV may compromise the police’s ability to detect and efficiently respond to incidents of concern. The size of many surveillance systems places a heavy burden on CCTV operators, with high camera-to-operator ratios meaning that many crime incidents occurring within sight of CCTV go undetected. Given the standard “differential response” policy of police dispatch CCTV may face significant processing delays, similar to what has been previously observed with calls for service.

This study analyzes CCTV’s relation to punishment certainty in Newark, NJ. CCTV detections and 9-1-1 calls for service occurring over a 3-year period are compared across dispatch processing times (i.e. the time between incident reporting and officer arrival on scene) and closure rates (i.e. whether the incident resulted in a police enforcement action). We further examine the frequency of CCTV activity and the impact of various factors on its (downward) linear trend.

Overall, calls for service exhibited significantly shorter dispatch processing times than CCTV detections. Calls for service also exhibited significantly shorter dispatch processing times for incidents classified as either high- or intermediate-priority. CCTV detections exhibited significantly lower process times for drug offenses, disorder offenses, and low-priority incidents.

CCTV detections experienced significantly higher closure rates than calls for service for 6 of the 8 crime types included in the analysis: overall crime, disorder offenses, drug offenses, other crime, high-priority incidents, and intermediate-priority incidents.

Surveillance activity steadily (and drastically) decreased over time. An average of 10.19 CCTV detections and 3.41 subsequent enforcement actions occurred per week over the total study period. The highest levels of activity were evident at the beginning of the CCTV program, when fewer cameras were in place. An average of 26.84 CCTV detections and 9.47 enforcement actions occurred during Phase 1 of the program, when only 11 cameras were installed. During phase 5 of the program—which increased the system size to 146 cameras—only 2.11 CCTV detections and 1.22 enforcement actions occurred per week.

Regression models identified a number of factors associated with decreased surveillance activity. Each new phase of camera installation reduced weekly levels of both CCTV detections and enforcement by approximately 47%. The weeks following police officer layoffs in November 2010 exhibited an approximately 86% reduction in weekly CCTV detections. The introduction of an acoustic gunshot detection system in Newark was associated with a 29% reduction in weekly CCTV detections.

Overall, this study supports the notion that CCTV increases certainty of punishment. Unfortunately, this increased effectiveness is largely negated by CCTV detections and enforcement becoming rare as the system expanded. Police may suffer from expanding CCTV systems absent a plan to maintain early levels of surveillance activity.

Saturation Foot Patrol in a High-Violence Area: A Quasi-Experimental Evaluation

Piza, E. and O’Hara, B. (2014)

Justice Quarterly, 31(4): 693-718

Research Brief
Version of Record (Open Access)

Key Takeaways

  • The Newark, NJ Police Department deployed a team of foot patrol officers in a 0.25 sq. mi. target area from 6pm-2am on a nightly basis
  • The evaluation measured the effect of the foot patrol intervention as compared to 2 separate control areas over a 1-year period
  • Total street violence as well as the disaggregate categories of aggravated assault and non-fatal shootings significantly decreased absent any displacement
  • Robbery suffered from substantial levels of both spatial and temporal displacement
  • Police should design large-scale foot patrol interventions in a manner that doesn’t allow robbers to easily gauge the scope of the intervention

Research Summary

In the summer of 2008 the Newark, NJ Police Department (NPD) commenced Operation Impact, a saturation foot patrol intervention modeled after an NYPD strategy of the same name. On a nightly basis, 12 police officers under the supervision of three supervisors patrolled a quarter-square-mile area of the city. Operation Impact patrols occurred from 6pm to 2am each night to coincide with the times which violence was highest in the target area. Foot-patrols were deployed as a deterrence mechanism to communicate increased risk of apprehension to potential offenders. NPD officials further expected the proactive activity of foot-patrol officers to disrupt street-level disorder and narcotics activity that may present opportunities for violence.

To measure the effect of the intervention, crime incidents were measured within four areas for the one year pre- and post-intervention periods: the target area, a catchment zone (meant to measure spatial displacement), and two control areas: the surrounding police precinct and a prospective target area, “Zone B,” identified during the analysis phase. Separate controls were incorporated to provide added perspective to any observed crime changes in the target area. Operation Impact’s intensive foot patrol approach is compared with tactics specific to the two control areas: intermittent, place-based enforcement in Zone B and “standard” responses to crime in the precinct. Any reductions would need to outperform both control areas to unequivocally validate the strategy.

Overall violence reduced by about 42% in the target area as compared to the precinct control area, with shootings and non-domestic aggravated assaults reducing over 60%.  Operation Impact sustained its positive effect when compared to Zone B, albeit to a lesser extent. Compared to Zone B, overall violence and aggravated assault reduced by 30% and 61%, respectively, in the target area. While the crime reduction reflects favorably upon Operation Impact, the displacement an lysis highlights certain challenges pertaining to robbery. During the non-operational period, robbery increased in the target area from 15 to 26 incidents (73%), which was more than double the increase observed in Zone B during the same time period. No other crime types increased in the target area during the non-operational period.

The spatial displacement analysis found that shootings experienced a modest diffusion of benefits effect, with significant decreases observed in the catchment area. Overall violence experienced an overall level of spatial displacement that was less than the reduction achieved within the target area. However, spatial displacement occurring during the non-operational period was larger than the target area reduction. A review of the raw counts shows robbery to be predominately responsible for this spatial displacement. Robbery increased 96% (from 25 to 49) in the catchment area. During the non-operational period robbery in the catchment area more than doubled (from 13 to 30) with an increase of 58% (from 12 to 19) observed in the catchment area during operational hours.

On a whole, this study offers support for foot patrol as a violence reduction tool, with Operation Impact reducing overall incidents of violence as well as the disaggregate categories of shootings and aggravated assault. However, while the place-based policing literature demonstrates displacement is far from inevitable, the robbery findings serve as a reminder that displacement can be a threat to crime prevention efforts. In light of these observations, additional effort may be necessary in order to avoid displacement if crime patterns include substantial levels of robbery.