October 24, 2013
The Riley County Police Department (RCPD) is a consolidated police agency, the only one in Kansas, and its mission is to reduce crime and improve the quality of life for the citizens it serves. It has been accredited by the Commission on Law Enforcement Accreditation (CALEA) since 1991. The RCPD currently has 205 employees, of which 107 are sworn police officers. The agency serves an area covering 614 square miles with a countywide population of nearly 76,000 residents. The City of Manhattan, with a population of approximately 53,000 residents, receives most of the police services in Riley County. Manhattan is also the home of Kansas State University, a Division I school with an enrollment of 23,000 students. Recent U.S. Census data shows that the Manhattan metro area was the 10th fastest growing in the country from 2010 to 2012. Much of the growth is a result of 1st Infantry Division troops returning to nearby Ft. Riley from deployment in Afghanistan, and significant future growth is likely to result from construction of the new National Bio and Agro-Defense Facility (NBAF) in Manhattan.
Despite the recent Manhattan area growth, the crime rate has steadily declined over the last four years, in large part due to the RCPD’s evidence-based crime reduction strategies that have focused on targeting repeat offenders and high-crime places. In early 2010, the RCPD implemented Operation Impact, a general smart policing initiative to reduce crime by focusing on areas susceptible to high crime rates. The Kansas State University research team of Williams and Kurtz (2011) assessed a segment of Operation Impact with regard to three categories of burglary – residential, commercial, and automotive. Four “impact zones,” each consisting of approximately 2,400 square feet, were selected for concentrated crime reduction and prevention efforts by patrol officers, while two additional high-crime patrol areas were held as control groups. Results of the assessment revealed a statistically significant reduction in total burglaries from 2009 to 2010 (year of implementation) within each of the four impact zones.
In late 2012, the RCPD decided to refine their strategy to incorporate the latest research regarding hot spot policing at “micro places,” or street segments no more than a city block in length. An experimental study by the Sacramento Police Department (Telep, Mitchell, & Weisburd, 2012) suggested that police officer presence in high-crime micro places for short periods of time (12-15 minutes) results in reduced crime and calls for service. The RCPD developed a similar strategy for Manhattan, named Initiative: Lase Point, and again enlisted the assistance of a research team from Kansa State University (Williams and Chernoff, 2013) to evaluate the strategy’s effectiveness from two perspectives. First, the existing literature has not addressed whether hot spot policing works in non-urban areas like Manhattan. Second, and more importantly, little has been done to assess behavioral practices of police officers within the hot spots. The RCPD study sought to provide practical and experimental insight to these two issues.
First, hot spots for the study were identified using by historical data and selecting street-length segments that experienced a relatively high number of crime incidents over the previous 12 months. The pre-analysis yielded 48 hot spots that were paired based upon similar attributes (apartment complex, trailer park, or residential area, etc.). None of the hot spots adjoined another in order to avoid treatment contamination. Each pair of addresses was randomized as either a “V” (officer visibility only) or “VA” (officer visibility and activity). The hot spots were divided into three groups of 16 (8 treatment and 8 control), generally attempting to maximize the time of day in which the areas were “hot” (more prone to criminal activity). Distribution was then aligned with the three-shift organization of the RCPD’s uniformed patrol division, assigning each shift eight pairs of addresses so that each hot spot could be visited at least once in a 24-hour period. The group of hot spots remained the same throughout the experimental period and was randomized each day by auto-generation; that is, visits were arranged in randomized order for that given day.
To reduce officer bias, hot spots were assigned to specific patrol officers by matched pairs, so that each officer visited both a V and VA hot spot. Further, supervisors were encouraged to involve as many patrol officers as feasible to minimize selection bias. Supervisors were instructed that each address was to be visited one time in the order that it appeared each day on the randomized list. Supervisors were given the freedom to make the visits happen as they saw fit; generally, a supervisor would assign the hot spots to individual officers prior to the start of the shift and the officers would conduct the visits when call load and operations allowed.
Officers visiting the V hot spots were instructed to visibly park in the area, remain there for 15 minutes, and to refrain from any proactivity unless required in the line of duty. Officers assigned to the VA hot spots were instructed to visibly park, get out of the car, and proceed with activities that included public contacts and order maintenance issues such as code enforcement, illegal parking, excessive noise, or alcohol-related violations. The experiment was launched on October 2, 2012, and continued through December 31, 2012. A total of 73 individual patrol officers participated in the micro hot spot visits, logging a total of approximately 825 hours. Assigned officers logged each visit into the department’s Records Management System (RMS). Incident and call data that fell in or within 25 feet of any study area were extracted from RMS and linked, using the ArcGIS 10 software system, to the hot spot locations with which they corresponded; visit data were then collated with these incident and call data. The department’s crime analysis unit tracked progress throughout the trials, providing periodic status reports. Variables tracked during each visit included shift, treatment designation, date, time in/out, and certain officer characteristics such as sex, age, and race. Outcome variables included counts of calls for service (including citizen- and officer-generated calls), Part I crimes, and Part II crimes/incidents.
At the conclusion of the study, crime incidents and call for service that took place in the 48 micro hot spots during the fourth quarter of 2009 through 2012 were analyzed in order to compare the same geographic areas during the same time period for three years prior to the experimental trials. The data were evaluated in part using the Wilcoxon Signed-Rank WSR) test, adjusting for ties (Higgins, 2004). The WSR test is a nonparametric, paired-comparison permutation statistical method designed to test whether or not a population of paired differences had a mean value statistically different from zero; it relies on the assumption that treatments were randomly assigned to the analytical units within each pair. The analysis compared matched pairs across time (i.e., before versus after the experimental period) and across space (i.e., “V” versus “VA” areas), testing whether or not hot spot policing had a noticeable effect in the community, and whether or not the effects were more or less effective depending on the kind of policing technique employed. In other words, does micro hot spot policing work in non-urban areas, such as Manhattan, Kansas? And does officer behavior within the hot spot matter?
Regarding the first question, the results of the study analysis demonstrated that hot spot policing at micro places, using the 15-minute treatment period established by Telep et al. (2012), resulted in a statistically significant decrease in calls for service and Part I and Part II crimes when comparing the same geographic areas over a four-year period. These results were not entirely surprising, as previous research has already established that hot spot policing does work to reduce crime, albeit research that focused on larger, urban areas. The second question, however, provided a more interesting answer. No significant difference was observed in either Part I or Part II crimes between treatment areas. In other words, there was little distinction in the decline of crime incidents between areas where officers were active and areas where officers merely established a visible presence. This strongly suggests that at least for this study, specific officer behavior (as defined here) did not result in significant differences in the reduction of crime incidents. Regardless, the study showed that Laser Point was successful in furthering the RCPD’s primary mission.
Aside from producing a verifiable reduction in crime and calls for service in the targeted areas, Laser Point produced additional benefits. During Operation Impact, the RCPD relied heavily upon high visibility traffic enforcement in high crime areas, resulting in some resistance from community members. Laser Point deployed officers more widely throughout Manhattan and Riley County than Operation Impact, and it relied more on neighborhood order maintenance and positive public contacts than traffic stops. In the end, Laser Point produced the same crime reduction benefits as Operation Impact but without any public perception of overly aggressive enforcement tactics. At a time when trust between the police and the community is so essential, deployment strategies like Laser Point suggest that the police can accomplish a mission of crime reduction without jeopardizing their legitimacy.
Laser Point also demonstrated that a smaller agency can undertake meaningful research and evaluation of new strategies. This particular experimental trial was developed and implemented entirely by the Riley County Police Department, with design and statistical consultation from researchers at Kansas State University. No external funding was used to support the project. Further, the trials and protocols were put into place by police supervisors and patrol officers, providing much-needed structure and consistency to the experiment. Because the agency has been proactive with previous evidence-based programs (e.g., Operation Impact), and because ongoing training addresses officer engagement, the transition and process went smoothly. The partnership between the RCPD and researchers at Kansas State University is also noteworthy, providing a relationship that produces valuable assessments that further our understanding of police practices and policies. During budgetary hard times, such partnerships are especially important.
Since the end of the experimental phase of Laser Point, RCPD has expanded its hot spot policing efforts at micro places. As above, analysis continues to incorporate historical problem places of crime. Additionally, recent crime incidents have been introduced into an algorithm with the historical data in an attempt to project a future risk for these types of crime at locations across the jurisdiction. This type of continued assessment and refinement of Laser Point will enable the RCPD to sustain the program as one of the primary crime reduction strategies in fulfilling its mission to the citizens of Manhattan and Riley County.
For more information on Initiative: Laser Point visit the Programs and Services section of RileyCountyPolice.org. Also view our press releases for more information on the award RCPD received for this research project.
Higgins, J.J. (2004). Statistics and Technology: Reflections on 35 Years of Change. Journal of
Modern Applied Statistical Methods, Vol. 3, No. 2 567-575.
Telep, C. W., Mitchell, R.J., & Weisburd, D. (2012). How Much Time Should the Police Spend at
Crime Hot Spots? Answers from a Police Agency Directed Randomized Field Trial in
Sacramento, California. Justice Quarterly (2012 Ahead-of-Print: 1-29).
Williams, L.S., and Chernoff, W. (2013). Making Gloves that Fit: Micro Hot Spot Initiatives in a
Non-urban Police Agency. Manhattan, KS.
Williams, L. S., & Kurtz, D. (2011). Initial Assessment Report: Riley County Police Department
Burglaries Project. Manhattan, KS.