Variability of Crime Clearance Among Police Agencies

Abstract

Average crime clearance rates have remained remarkably stable in the United States since the 1980s, despite many advances in investigative technologies or fluctuations in crime. Taking these average trends at face value, some have suggested that this stability indicates that police departments can do little to alter their clearance rates. However, in this study, we find that the average trends mask substantial long-term variation in crime clearance among police agencies. Using group-based trajectory modeling, we test whether large U.S. police departments have reported uniquely different long-term clearance rate trends from 1981 to 2013 and what organizational factors might contribute to different trends. As we discuss, this method has attractive qualities that provide for a more rigorous analysis compared with past comparative work. Our results show diverse levels and patterns of clearance both within individual crime types and across multiple crime types that appear to covary with organizational factors. We explain how finite mixture modeling can advance both quantitative and qualitative research by identifying departmental differences in performance for further study.

Clearing homicides

“…In the first phase of our analysis, we carried out a trajectory analysis of case clearances for homicide, robbery, aggravated assault, and burglary to understand the long‐term clearance rate trends of large U.S. agencies. We report the findings from that analysis in Scott, Wellford, Lum, and Vovak (; see also Lum, Wellford, Scott, & Vovak, ). Using data derived from the “Offenses Known and Clearances by Arrest” summary data as reported to the Federal Bureau of Investigation (FBI) Uniform Crime Reports (UCR) program, we created yearly clearance rate trends from 1981 to 2013 for the 100 largest agencies in the United States for homicide, robbery, aggravated assault, and burglary.…”

“…The three approaches used to rank‐order the agencies from highest to lowest performing with regard to clearance rate trends across the four serious crime types were (1) assigning ordinal labels to trajectory groupings from lowest to highest long‐term trajectory; (2) assigning probability weights from the trajectory analysis itself to create the ranks; and (3) using multitrajectory modeling techniques to estimate trajectory groups for multiple groups simultaneously (see Scott et al., ).…”

“…The choice of eight agencies was estimated given the available funding for this project.7 The three approaches used to rank-order the agencies from highest to lowest performing with regard to clearance rate trends across the four serious crime types were (1) assigning ordinal labels to trajectory groupings from lowest to highest long-term trajectory; (2) assigning probability weights from the trajectory analysis itself to create the ranks; and (3) using multitrajectory modeling techniques to estimate trajectory groups for multiple groups simultaneously (seeScott et al, 2019).8 By selecting agencies at the extreme, we are more likely to identify agency-level differences than if we had taken a random sample of the same size of all agencies. We understand that the other trajectories not sampled in this study might be explained by other factors.…”

The Effects of Body-Worn Cameras on Police Efficiency: A Study of Local Police Agencies in the US

“…Thus, the literature review focuses on the strand of research that more closely approximates the study of efficiency as an outcome, although no studies to date have used a measure of police efficiency as their primary outcome of interest. For a thorough review of available evaluations and research on BWCs, see Lum et al (2019).…”

“…The number of research studies focusing on this proxy of police efficiency, however, is relatively small; it accounts for 6% of all the published research on BWCs to date (Lum et al, 2019), and the results are mixed. Studies have examined the impact of BWCs using the gold standard for evaluations (RCTs) or quasi-experimental approaches, and "before and after" approaches as well as qualitative analyses to support their quantitative findings.…”