Article 0

The killing of George Floyd in Minneapolis USA in May 2020 raised questions about the impartiality of police officers when dealing with the public and particularly those from black and ethnic minorities.  Six years earlier, in 2014, a team of researchers from Stanford University investigated exactly this issue by using footage from police body-worn cameras to analyse the language they used during routine traffic stops in the multiethnic city of Oakland in California. 

The researchers transcribed footage from 981 traffic stops of both black and white drivers, which were conducted by 245 different officers during a period of one month. Participants in the study were given transcripts of these interactions without knowing the race, age or sex of the drivers. They were then asked to rate the respect shown by the officers through placing their language on a gradient, showing how respectful, polite, friendly and formal they were. The research team used a model based on linguistic theories of respect where respectful language includes apologising, being grateful, expressing concern for the other person and softening commands to reduce confrontation. 

The results demonstrate that, although officers used the same levels of formality for both black and white drivers, they were rated as significantly less respectful, polite or friendly to black drivers than they were to whites.  Even after controlling for the severity of the traffic offence, the length and outcome of the stop and the race of the officers themselves, interactions with white drivers were consistently more respectful. In fact, 57% of white drivers were more likely to hear an officer say one of the most respectful phrases (e.g. “sir”, “thank you”) in the transcripts whereas black drivers were 61% more likely to hear one of the least respectful (e.g. “hands on the wheel”, or use of first name). 

The researchers conclude that the racial disparities in their study are clear, however the causes of them are not.  They write that these disparities could have far-reaching effects as personal interactions with the police build a community’s opinion about them and ultimately lead to a relationship of trust or distrust. They suggest that future research could expand body camera footage beyond just text to audio features such as intonation and video features such as facial expression, to try and investigate how interactions progress and sometimes break down.  This could be invaluable in informing police officer training and to establish better relationships with the communities they serve. 

Voigt, Rob, Nicholas Camp, Vinodkumar Prabhakaran, William Hamilton, Rebecca Hetey, Camilla Griffiths, David Jurgens, Dan Jurafsky, and Jennifer Eberhardt. 2017. Language from police body camera footage shows racial disparities in officer respect. PNAS 114 (25): 6521-6526.

https://www.pnas.org/content/114/25/6521

This summary was written by Gemma Stoyle