Linking body worn camera activation with complaints: The promise of metadata

With the introduction of body worn cameras, new data types have emerged, including activation metadata, which is information on the extent of the implementation of body worn cameras, by whom, and under what conditions. In this paper, we propose an avenue of methodological interest: linking activation metadata with police-recorded behaviour. We take the case of complaints to examine the use of these data for investigations of officers’ misconduct. We used an observational approach, analysing 1.73 million body worn camera activations by more than 3900 frontline officers, juxtaposed with professional conduct data. We find a heterogeneous distribution of implementation, despite a ‘blanket’ body worn camera policy requiring activation. Specifically, distinct types of officers disproportionally under-activate body worn cameras. We show that body worn cameras activation rates are inversely correlated with complaints, although minimally when observed at the population level. We discuss the use of metadata for research as well as for policy.