According to new research from Stanford, the first systematic analysis of footage from police body cameras showed that officers consistently use less respectful language with blacks that with whites in everyday traffic stops. According to the study conducted by PNAS (Proceedings of the National Academy of Science), “Such disparities in common, everyday interactions between police and the communities they serve have important implications for procedural justice and the building of police–community trust.”
It is important to note that the result was obtained even after controlling for the race of the officer, the severity of the violation, the stop location, and the outcome of the situation.
According to the study published on June 5, the study presents “a systematic analysis of officer body-worn camera footage, using computational linguistic techniques…we develop computational linguistic methods that extract levels of respect automatically from transcripts, informed by a thin-slicing study of participant ratings of officer utterances.”
These techniques are designed to automatically measure the respect levels displayed by officers. Officer body camera footage, can be utilized as a tremendous source of data rather than just historical evidence. The study also opens the door for the potential development of language-based tools to study and potentially improve police-community communications.
Jennifer Eberhardt, a professor of psychology and co-author of the study, said: “Our findings highlight that, on the whole, police interactions with black community members are more fraught than their interactions with white community members.”
Nonetheless, Eberhardt acknowledges the challenges in making accurate conclusions from hundreds of hours of footage.
However, none of these officers were swearing, all were “well behaved” as described by Dan Jurafsky, a professor of linguistics as well as a co-author, “But the many small differences in how they spoke with community members added up to pervasive racial disparities.”
“The fact that we now have the technology and methods to show these patterns is a huge advance for behavioral science, computer science and the policing industry…Police departments can use these tools not only to diagnose problems in police-community relations but also to develop solutions,” said Rob Voight, lead author of the study and a doctoral student of linguistics at Stanford.
A new artificial intelligence technique for measuring levels of respect in officers’ language was developed by a multidisciplinary team from Stanford’s psychology, linguistics, and computer science departments. The technique was then applied to 981 traffic stops by the Oakland Police Department in a month.
The results showed that whites were 57 percent more likely to hear respectful utterances, expressions of gratitude and even apologies. However, blacks were 61 percent more likely to hear less respectful utterances or commands, or even be referred to by informal titles such as “dude” or “bro.”
The study consisted of three phases of research, accounting for time constraints and biases of the researchers.
However, the researchers do not point fingers of bias directly at the officers, accusing them of bias or racism. “Our findings are not proof of bias or wrongdoing on the part of individual officers. Many factors could drive racial disparities in respectful speech,” acknowledged Eberhardt.