The Kansas City Preventive Patrol Experiment

This landmark experiment found that traditional routine patrol in marked police cars does not appear to affect the level of crime. Nor does it affect the public’s feeling of security. The experiment demonstrated that urban police departments can successfully test patrol deployment strategies, and that they can manipulate patrol resources without jeopardizing public safety.

Patrol is considered the backbone of police work. Billions of dollars are spent each year in the United States to maintain and operate uniformed and often superbly equipped patrol forces. The assumption underlying such deployment has been that the presence or potential presence of officers patrolling the streets in marked police cars deters people from committing crime.

But the validity of this assumption had never been scientifically tested. And so, in 1972, with funding and technical assistance from the Police Foundation, the Kansas City Police launched a comprehensive, scientifically rigorous experiment to test the effects of police patrol on crime.

The Experiment

The experiment began in October 1972 and continued through 1973; it was administered by the Kansas City Police Department and evaluated by the Police Foundation.

Patrols were varied within 15 police beats. Routine preventive patrol was eliminated in five beats, labeled “reactive” beats (meaning officers entered these areas only in response to calls from residents). Normal, routine patrol was maintained in five “control” beats. In five “proactive” beats, patrol was intensified by two to three times the norm.

The experiment asked the following questions:

  • Would citizens notice changes in the level of police patrol?
  • Would different levels of visible police patrol affect recorded crime or the outcome of victim surveys?
  • Would citizen fear of crime and attendant behavior change as a result of differing patrol levels?
  • Would their degree of satisfaction with police change?

Information was gathered from victimization surveys, reported crime rates, arrest data, a survey of local businesses, attitudinal surveys, and trained observers who monitored police-citizen interaction.

Major Findings

Interestingly, citizens did not notice the difference when the level of patrol was changed. What is more, increasing or decreasing the level of police patrol had no significant effect on resident and commercial burglaries, auto thefts, larcenies involving auto accessories, robberies, or vandalism–crimes traditionally considered to be prevented by random, highly visible police patrol.

The rate at which crimes were reported to the police did not differ in any important or consistent way across the experimental beats. Citizen fear of crime was not affected by different levels of patrol. Nor was citizen satisfaction with police.

“Ride-alongs” by observers during the experiment also revealed that 60 percent of the time spent by a Kansas City patrol officer typically was noncommitted. In other words, officers spent a considerable amount of time waiting to respond to calls for service. And they spent about as much time on non-police related activities as they did on police-related mobile patrol.


The findings do not prove per se that a highly visible police presence has no impact on crime in selected circumstances. What they do suggest, however, is that routine preventive patrol in marked police cars has little value in preventing crime or making citizens feel safe.

The overall implication is that resources ordinarily allocated to preventive patrol could safely be devoted to other, perhaps more productive, crime control strategies. More specifically, the results indicate that police deployment strategies could be based on targeted crime prevention and service goals rather than on routine preventive patrol.

It is important to note that this experiment was conceived and executed by a local police department with technical help from outside researchers. The experiment demonstrates that, with the right kind of leadership and help, urban police departments can test new approaches to patrol. And they can use their patrol resources to conduct such experiments without jeopardizing public safety.

The full report can be found here.