Does More Training Mitigate Disparities in Police Use of Force? Quasi-Experimental Evidence from New Linked Data

Awarded Scholars:
Matthew Ross, Claremont Graduate University
CarlyWill Sloan, Claremont Graduate University
Project Date:
Jun 2021
Award Amount:

Police training is considered a key approach to reducing police violence and disparate treatment. However, few have studied the impact of training on the incidence of force and there is little evidence that such interventions will mitigate racial and ethnic disparities. Economists Matthew Ross and CarlyWill Sloan will examine police training programs and their impact on the use and escalation of force and the extent to which trainings mitigate racial and ethnic disparities. They will build and analyze a dataset that links course-level training records for all Dallas Police Department officers with the universe of nearly three million 911 calls and subsequent use of force and arrest reports from 2014-18. The 911 data contain dispatcher notes describing the incident and force applications by officers, as well as the sequence of events. Ross and Sloan will focus on three required courses mandated by the Texas Legislature for all officers including: (1) de-escalation; (2) crisis intervention, and (3) use-of-force training. All officers must take 40 credit hours of crisis-intervention training and eight of de-escalation training every four years until they achieve advanced proficiency (after 8 to 21 years of service) and 16 credit hours of use-of-force training until they achieve intermediate proficiency (after 3 to 9 years of service). The training data include about 650,000 individual course enrollments for all officers employed at any point from 2006-19, which overlaps with the sample of 911 calls. Information is available on the date, time, and location of each training course, the course name, credit hours, and instructor. The Texas Commission on Law Enforcement and the Texas Law Enforcement Education Agency have provided the investigators with course syllabi and instructional materials for each of three types of courses and for several other required courses. Ross and Sloan will: (1) link 911 calls with use-of-force incidents, dispatcher notes, arrests, officer training, and officer administrative records; (2) use natural language processing to create escalation variables using the text data provided in the dispatcher note; (3) use cluster analysis techniques to construct measures describing the escalation of force during an incident.


RSF: The Russell Sage Foundation Journal of the Social Sciences is a peer-reviewed, open-access journal of original empirical research articles by both established and emerging scholars.


The Russell Sage Foundation offers grants and positions in our Visiting Scholars program for research.


Join our mailing list for email updates.