Measuring and Analyzing Community- and Officer-Initiated Activity: Developing a National Program

RTI International, the National Policing Institute, and AH Datalytics, along with agency and stakeholder partners, will conduct research and development activities to provide actionable information to the Bureau of Justice Statistics (BJS) as they seek to develop a nationally representative, standardized set of indicators from law enforcement Computer-Aided Dispatch (CAD) systems to understand law enforcement activities and workloads, assess community needs, and identify opportunities to improve public safety holistically, particularly as jurisdictions seek to understand how best to respond to community needs. CAD data are the most relevant and widely available data source to understand the scope of law enforcement officer activity to both community-initiated calls for service and officer-initiated activities. While the United States has historically had two national data sources about crime through BJS’ National Crime Victimization Survey and the FBI’s Uniform Crime Reporting Program, it lacks a nationally representative data source that can provide reliable information about criminal and non-criminal police–community interactions that occur before a crime incident is document by law enforcement. This opportunity comes with key challenges, including a lack of data standardization across CAD systems and jurisdictions, limited data quality and completeness checks, and uncertainty around data fitness for statistical purposes. 

Drawing from longstanding thought leadership and technical knowledge on using CAD data for research, analysis, and practical applications, this team will complete three key objectives to provide BJS with information to address these challenges so that sound decisions can be made about designing and implementing a national program. In Objective 1, they will involve practitioners and researchers to identify key CAD metrics. Five initial agencies will also be engaged to obtain sample CAD data, which will be evaluated to determine available, consistent CAD data elements, which will then be mapped and standardized using a machine learning process with human review. For Objective 2, they will develop a proof of concept (POC) database using the specifications developed in Objective 1. The standardized code and rules will be translated into a programmatic process, which will be tested with CAD data representing an additional 15–20 jurisdictions using open-source CAD data. In Objective 3, they will assess the POC database’s ability to generate estimates of CAD events and will develop a road map that describes how a national program can be scaled up, including the number and types of agencies that would need to be included for nationally representative estimates.