Measurement of Potential Over- and Under-policing in Communities

Abstract Over- and under-policing of neighbourhoods can undermine public trust and confidence in the police as well as the broader justice process. This study reports on attempts to operationalize and test a spatial indicator of potential over- and under-policing, where over-policing is defined as a level of police presence at a particular location that is greater-than-expected, given the level of public demand for police services, current police enforcement strategy, and community preference regarding police activity. Automated Vehicle Locator data and Computer-aided dispatch logs from the Seattle Police Department, as well as data drawn from community-based surveys, are modelled using a Geographic Information System. The model uses 2-week data windows to provide timely and actionable information that can be rendered for decision makers in a CompStat style accountability and management forum. Such an approach has potential utility for police management, as well as for community engagement and reform efforts aimed at addressing the problem of over-policing.