Date of Award

Summer 8-1-2020

Degree Type


Degree Name

Doctor of Philosophy (PhD)


Criminal Justice

First Advisor

Eric Sevigny

Second Advisor

Richard Rosenfeld

Third Advisor

William Sabol

Fourth Advisor

Richard Wright


This dissertation consists of three papers examining whether and to what degree educational capital (e.g., college-educated officers and education policy) in law enforcement influence police-citizen violence, misconduct, and crime.

Paper 1 examines the effects of officer education on police-citizen violence. Using the Police Stress and Domestic Violence in Police Families in Baltimore, Maryland data, we employ a doubly robust propensity score design to compare outcomes among 1,104 Baltimore city police officers. We also apply multiple imputation to address missing data and include measures capturing officer work attitudes and stress levels. We find that having a bachelor’s degree is associated with a lower probability of an on-duty shooting relative to the counterfactual condition of high school education only. Results also indicate that college-educated officers are just as likely as non-college educated officers to make violent arrests and encounter physical altercations.

Paper 2 analyzes the relationship between the proportion of bachelor-degreed

officers in agencies and misconduct complaints. Using a national sample of 1,023 U.S. law enforcement agencies from the Police Use of Force: Official Reports, Citizen Complaints, and Legal Consequences Project data, this study employs design-based analysis and multiple imputation of missing data with a doubly robust propensity score model applied to a multivalued treatment. We observed that large city departments in the medium and high education categories generated more abuse of authority and overall misconduct complaints than departments with low education levels.

Chapter 3 examines the mediating effects of arrests on the departmental education requirements and crime association. We integrate entropy balance weighting procedures with a novel mediation approach to identify the direct and indirect effects of agency education policy on crime in 1,543 U.S. cities using multiple data sources: Law Enforcement Management and Administrative Statistics (LEMAS) survey, U.S. Census Bureau’s American Community Survey (ACS), and FBI’s Uniform Crime Reports (UCR). We find that police arrest productivity does not significantly mediate the impact of education policy on crime outcomes; however, significant total effects were detected. For violent crime, college requirements were associated with significant reductions in aggravated assaults. For property crime, cities adopting a college requirement for police experienced fewer burglaries.

Available for download on Sunday, July 24, 2022