Information Technology on the Beat: The Impacts of Body-Worn Camera and Facial Recognition Technology on Public Safety

Abstract

Body-worn cameras (BWCs) have transformed the criminal justice system by facilitating information gathering and thereby enhancing public safety. Despite the promise of this technology, there is also much skepticism about its efficacy. This study examines the effect of BWC implementation on the likelihood of a suspect being arrested and on the overall crime rate. In addition, we delve into whether and how the BWC, an information gathering tool, interacts with facial recognition technology, a data analysis tool. To identify the impacts of BWC implementation, we exploit a quasi-experimental setting, in combination with a matching technique, by using a federal grant program for BWC implementation as a natural shock. Using granular crime incident–level data during the period 2010–2016, we find that the adoption of BWCs is positively associated with the probability of apprehension under its jurisdiction, especially on-view (warrantless) arrests based on a probable cause. Furthermore, machine learning–powered facial recognition technology appears to augment the effectiveness of BWC in apprehension. Using police agency–level data, the implementation of BWCs is found to be negatively associated with overall crime rates. This study sheds new light on the societal value of IT and machine learning in the public sector, beyond its business value, providing meaningful implications for policymakers and researchers.