Predictive Policing Symposium: The Future of Prediction in Criminal Justice
Defining Terms and Introducing Issues
Theron Bowman, Ph.D., Chief of Police, Arlington, Texas
Dr. John Morgan opened the discussion by introducing a working definition: “Predictive policing refers to any policing strategy or tactic that develops and uses information and advanced analysis to inform forward-thinking crime prevention.” This definition and the five elements of predictive policing that Morgan introduced appear in a Harvard Executive Session paper being drafted by Dr. Morgan, Chief Bill Bratton and Lieutenant Sean Malinowski.
Ultimately, predictive policing is intended as a framework to advance strategies like community policing, problem-oriented policing, intelligence-led policing and hot spots policing.
Dr. Theron Bowman provided a working example of predictive policing from the Arlington (Texas) Police Department. Chief Bowman explained how his department uses data on residential burglaries to identify hot spots, and then compares these locations to places with code violations in the city. With analysis, the data serve to monitor the relationship between things like neighborhood physical deterioration and crime.
Given the social disorganization traditionally associated with a large amount of code violations, Arlington has developed a formula to help identify characteristics of a “fragile neighborhood.” As a result, the police department and other city agencies now apply resources more efficiently to these fragile neighborhoods, ultimately preventing crime.
Jeremy Crump, director of strategy for the National Policing Improvement Agency (the British sister organization of NIJ), discussed police prediction in the United Kingdom. He said that although the term “predictive policing” has not yet been used in the U.K., the fundamentals are drivers of change. For example, the U.K. has been using a national database (Police National Computer) for 35 years.
Crump stressed the importance of adapting to changing conditions and focusing less on the hierarchy that exists within many agencies. For predictive policing to make a difference, departments need to be willing to explore that change.
Dr. Craig Uchida discussed the role of research and theory in predictive policing. He emphasized the need for theories to drive the information being analyzed. Uchida stated that prevention theories are already available; this information must now be integrated with data that is also available. He discussed how situational crime prevention, social disorganization and collective efficacy theories may include predictive analytics, and suggested that they should be used in guiding policing strategies and tactics.
