Reducing crime is a constant concern of law enforcement and community leaders. Police strategies for reducing crime rely heavily on deterrence, in the form of police patrols (reviewed by Koper, 1995, p. 649-650). Research has shown that a police presence reminds offenders and potential offenders of the certainty of punishment, which is a more effective deterrent than the promised severity of a punishment. The findings from early studies on the effectiveness of police patrols as a crime deterrent were mixed, but with publication of a well-controlled Kansas City study in 1986 the debate moved on to what factors influence the deterrence effect.
Of the variables that have been found to influence criminal activity, geographic location stands out (Koper, 1995, p. 652). A study done in Minneapolis revealed that just 3.3% of the city's addresses and intersections accounted for over 50% of the requests for police help. These 'hotspots' for criminal activity included those where serious crimes occurred, such as robbery, criminal sexual assault, and auto theft. By focusing preventive policing efforts on hotspots, deterrence would be predicted to have the greatest impact.
Another variable suspected of influencing the deterrence effect is disorder, in the form of behavior and the physical appearance of the location (Koper, 1995, p. 651). For example, broken windows, boarded up buildings, and graffiti has been suggested to foster the perception of a lack of safety and high crime. Such locations are believed to foster certain behaviors related to disordered locations, such as vagrancy, panhandling, vandalism, drunkenness, drug use, and prostitution. When researchers examined robbery data for several neighborhoods, they found that disorder provided an indirect link between crime prevalence and economic/social decay.
The effectiveness of a police presence in reducing crime has been well established, but the variables that influence the magnitude of the effect are still being worked out. Of primary concern is how long the effect lasts after the police have left the area (Koper, 1995, p. 658). This 'residual' deterrence effect is defined as criminal activity remaining below normal levels after police have left the area. An initial analysis of the Minneapolis data revealed that disorder and criminal behavior decreased by 25 and 65%, respectively, immediately after the police had left the area.
To further define the parameters of residual deterrence, Koper (1995) analyzed the Minneapolis data to uncover what length of time a police presence had to persist in order to provide the maximum deterrence effect. Excluded from the data were instances when a disturbance elicited a police response (Koper, 1995, p. 661). The average length of time that a police presence had to persist to produce the maximum residual deterrence effect was 14 to 15 minutes (Koper, 1995, p. 664). The effectiveness is quite substantial; reducing the probability of a disorder occurring to about 4%, 30 minutes after the police had left the area.
If the time the police remained in the area was grouped into 1 to 5 minutes, 6 to 10 minutes, 11 to 15 minutes, or 16-20 minutes, only a stay of 11 to 15 minutes produced a significant (p < 0.01) residual deterrence effect (Koper, 1995, p. 663). The magnitude of the effect was a 388% increase over the residual deterrence effect of a police drive-by (zero minutes). However, a 1 to 5-minute police presence produced a worse outcome compared to drive-bys and the other time groups were not significantly different.
An interesting result from this study is that drive-bys alone are very effective in producing a residual deterrence effect. For example, 10 minutes after a drive-by the chances that a disorder would occur is just 6.5%, and at 30 minutes, 16%. Drive-bys are more effective than a police presence lasting 1 to 5 minutes and similarly effective to police presences lasting 6 to 10 or 16 to 20 minutes.
Koper's (1995) research reveals how residual deterrence, caused by a police presence unrelated to a disturbance, can reduce the prevalence of disorder in crime hotspots. Disorder is used as an outcome measure, in part because the number of crimes in the data was too few to provide sufficient statistical power, and because research findings have suggested that disorder provides a link between social/economic decay and criminal activity. Despite this limitation, the above findings reveal that drive-bys and a police presence lasting between 11 and 15 minutes provides a significant residual deterrence effect lasting at least 30 minutes. Whether the crime was displaced to another location could not be determined from this analysis.
Based on the conclusions drawn from a recent study examining the nature of policing research, in terms of its rigor and predictive potential, the focusing of police patrols on hotspots is one of the best-validated, evidence-based policing strategies in existence (Lunn, Koper, and Telep, 2011, p. 6-7). This conclusion is consistent with findings that suggest policing strategies are most effective when focused, proactive, and place-based. The focus of the hotspot research study was disorder deterrence caused by a short-term police presence. It was proactive because it relied on historical crime and disorder data to assess strategies designed to prevent future crime. It was place-based because it focused on intersections or addresses as disorder/crime hotspots.
Using the policing strategies recommended by these studies, the aftermath of the media storm surrounding the Hayley Scott murder (Sparrow, 2009) should be handled in a focused and proactive manner. Although the Hayley Scott had a history of complaints about a stalker, no definitive descriptions were reported that could provide the investigators with a reasonable lead. The murder occurred at an interstate rest stop and could therefore have been a random event by someone traveling through the area. In support of this possibility, Hayley Scott never reported any death threats arising from the stalking activity. The 61 women, who reported that they suspected they were being stalked as well, filed these reports only after the media storm about the unsolved murder. In addition, as a group they could provide no specifics about their stalkers. In the absence of any solid leads, Lunn and colleagues (2011) would probably recommend that the Heron Police Department focus on historical hotspots for sexual assault and murder. For example, the interstate rest stop would probably qualify as a hotspot by virtue of the Hayley Scott murder and random drive-bys or a police presence lasting 14 to 15 minutes would be recommended.
Contrary to Major Lucius' hesitancy to use CompStat without any solid leads concerning the current stalking complaints in Heron City (Sparrow, 2009, p. 6), CompStat could be used to generate hotspot predictions based on the historical data for sexual assault, stalking, and stranger-on-stranger murder. The 143 calls from the 61 women could then be crosschecked with this hotspot map to see if any of them are close to these locations. These locations could then be targeted with random police drive-bys or a 15-minute police presence.
The same approach could be taken regarding the upsurge in luxury car thefts in Heron City (Sparrow, 2009). Like most cities, luxury cars would probably tend to be aggregated in specific parts of the city, but Major Lucius reported that the locations of the thefts seem to be randomly dispersed across the various precincts (Sparrow, 2009, p. 9); however, this does not imply that they are randomly dispersed within precincts. Major Lucius also mentions the failure of the automatic license plate reader system to produce a significant advantage in locating stolen cars. Again, the CompStat system could generate a map of car theft hotspots, maybe even luxury car hotspots, based on historical data (Taylor, Koper, and Woods, 2012). The position of the 18 ALPR cameras could then be moved to selected hotspots, if they are not already there. In addition, a specialized vehicle theft unit could be formed that manually checks these hotspots in the hopes of deterring car theft (Taylor, Koper, and Woods, 2012, p. 40).
Koper, C. (1995). Just enough police presence: Reducing crime and disorderly behavior by optimizing patrol time in crime hotspots. Justice Quarterly, 12, 649-672.
Lunn, Cynthia, Koper, Christopoher S., and Telep, Cody W. (2011). The evidence-based policing matrix. Journal of Experimental Criminology, 7, 3-26.
Sparrow, Malcolm K (2009). One week in Heron City: A case study. Program in Criminal Justice Policy and Management, Harvard Kennedy School. Retrieved 6 May, 2012 from http://www.ncjrs.gov/pdffiles1/nij/227664.pdf.
Taylor, Bruce, Koper, Christopher S., and Woods, Daniel. (2012). Combating vehicle theft in Arizona: A randomized experiment with license plate recognition technology. Criminal Justice Review, 37, 24-50.
Darrel Stephens' (2010) offers a what may be an uncommon, but well informed perspective on how the police (practitioners) and researchers can come together to try and improve policing methods. In mentioning MacDonald's "A dialogue of the Deaf," Stephens emphasizes how far apart practitioners and researchers are in the perspective each have towards policing strategies. Much of the lack of understanding between these two groups, according to Stephens, is not due to disagreements about whether policing methods…