The Kansas City Preventive Patrol Experiment

***

Advanced Research Methods

Week 2

Discussion 1

***

The Kansas City Preventive Patrol Experiment

The Kansas City Police Department undertook an experiment to determine if changes in preventive patrolling had an effect on police services (such as reducing crime levels, maintaining a community free of the fear of crime, and even the publics perception of the police). The experiment was conceived based on a shortage of resources, and the “concern was that any serious attempt to deal with priority problems would be confounded by the need to maintain established levels of routine patrol. (Kelling, Pate, Dieckman, and Brown, 1974, p. 40).

To determine if preventative patrolling had an an effect on services, the researchers basically conducted two experiments; the first, in which no preventative patrolling was conducted (this method named as “reactive”), and the second in which 2 to 3 times the number of preventative patrols were conducted (this method named as “proactive”). In both cases, the independent variable, the level of preventative patrolling, was compared to the dependent variable, the effect on police services. To establish a controlled experiment, the area was divided into three test areas based on the “basis of crime data, number of calls for service, ethnic composition, median income and transiency of population” (Kelling et al, 1974, p.7) This was intended to remove other possible variables from affecting the cause and effect relationship between the level of preventative patrolling conducted and the level of police service. Three areas were designated so that both the “reactive” and “proactive” methods could be compared against a control group. Units from the “reactive” method were used to increase patrols in areas designated in the “proactive” method. The same officers that had patrolled the area previously were used in the experiment to eliminate additional variables from affecting the cause and effect relationship. In addition, the areas under experiment were layed out in order to maintain response times. Although the experiment was initially flawed by the failure to maintain the conditions set for the experiment, these issues were identified and corrected; the experiment continued for 1 year.

The researchers measured the effects on police services in the following ways: a victimization survey, crime reported to the police, the arrest rate, a commercial survey, traffic data, encounter surveys from citizens, officers, and observers, officer surveys on noncomitted time, dispatch data, response time surveys from observers and citizens, and finally, a data analysis of officer activity. The researchers concluded that there were no statistically significant difference in any of the areas of police services between the control group and either the “reactive” group or the “proactive” group. There was one exception in the case of “other sex crimes”, but the researchers concluded that “not traditionally considered to be responsive to routine preventive patrol” (Kelling et al, 1974, p.16),. The researchers also found that there was no spillover or displacement effect.

The Kansas City Preventive Patrol Experiment and it's conclusions affected criminal justice policy in several ways. The primary effect was that “it suggested the implementation of targeted crime prevention strategies” (Avdija, 2008, para. 18) Carter suggests that the“implications for community policing were important, not the least of which was the fact that officers could free up time from patrol therefore using that time more efficiently through problem solving” (2000, p.3). The experiment can be said to have provided basis for both the community-oriented policing and the problem-oriented policing concepts.

References

Avdija, A. (2008, July-December). Evidence-based policing: A comparative analysis of eight experimental studies focused in the area of targeted policing. International Journal of Criminal Justice Sciences, Vol. 3. Iss. 2., Retrieved July 11, 2014 from http://www.sascv.org/ijcjs/avdi.html

Carter, D. (2000). Reflections on the move to community policing. Regional Community Policing Training Institue at Wichita State University. Retrieved July 11, 2014 from http://webs.wichita.edu/depttools/depttoolsmemberfiles/rcpi/Policy%20Papers/Reflections%20on%20Comm%20Pol.pdf

Kelling, G., Pate, T., Dieckman, D.,and Brown, C. (1974) The Kansas City Preventive Patrol Experiment. Police Foundation. Retrieved July 8, 2014 from http://www.policefoundation.org/content/kansas-city-preventive-patrol-experiment-0

***
Followup
***

I have a pretty good article floating around my drive on police work as an "art" versus a "science". (I'll see if i can dig it up, but I'm still way behind on organization)
Essentially, the point was that there can be a mutual distrust between academia and LE professionals.  Some academics look down on LEO's and don't trust field experience/department culturization, while street cops can have a distrust in "eggheads", whose ideas can run from unfeasible to purely destructive (see Cloward-Piven).
That isn't to say everyone in those respective fields feels that way, but the author felt there were enough people on both sided to make departments resistant to change led by academia.
If you also take into account organizational biases, it would be easy to say that any organization will be resistant to change.

 ***
One thing that would help would be for academics to mitigate any possible damage their research might cause; in the Kansas City Preventative Policing experiment, for example, both the department and the researchers watched for skyrocketing rates of crime, in which case they prepared to abandon the experiment

***

This comes back to the scarcity of resources issue; by cutting back the number of hours used in preventative patrolling, departments can target resources more efficiently, so I agree with you, this was definitely a success.

***
The more data you have, the more accurate your statistics are going to be (side note - I was surprised to find out how well the bell curve represented reality in my statistics class).  By extending the time the experiment ran, the researchers were able to gather more data.

In addition, in covering a year, the experiment was also able to take into account seasonal variations (in summer, you'll have more unsupervised vagrants and more juveniles will be out and about)

oops, please swap vagrants and juveniles in the above!

***
[Withheld] brought up the targeting of "hot spots" earlier; this would be the best way to start allocation of resources. "In Minneapolis, for instance, only 3 percent of the city’s addresses accounted for 50 percent of calls for service to the police" (Braga, 2008, p.6)
The tactics of how to apply the resources is a another matter of discussion.  Do you use surveillance teams, station additional reactive teams in the area, or take another tact?

Braga, A. (2008).Crime Prevention Research Review No.2: Police Enforcement Strategies to Prevent Crime in Hot Spot Areas. Washington, D.C.: U.S. Department of Justice Office of Community Oriented Policing Services,. Retrieved July 12, 2014 from
http://www.cops.usdoj.gov/Publications/e040825133-web.pdf