People of color face disparities in how they are evaluated from an early age. As one prime example, Black and Latino boys are suspended or expelled from school at 3.2 times and 1.3 times the rate of White boys. And those inequities in treatment persist into adult life when it comes to treatment by police, chances of getting a job or a promotion, or receiving the highest quality medical care.
“Whether it’s a police officer making a decision to stop a particular suspect, a doctor making decisions about how much pain a patient is in, or a hiring manager making decisions about why someone would be late to a job interview, anytime there’s a behavior evaluation taking place, there’s a possibility of bias on the part of the observer based on racial or ethnic background,” says Jayanti Owens, a sociologist and assistant professor of organizational behavior at Yale SOM.
Ownens’ prior research and that by other scholars has shown that Black students are more likely than White students to be punished for the same behavior. Yet one limitation with much of this prior work, says Owens, is that it is often based on subjective reports of student behavior on the part of teachers.
“When asking teachers in their school settings to rate the frequency of misbehavior by students from different backgrounds, I had to take for granted that these reports did not themselves contain racial bias,” she says. “In reality, I knew that it was possible that teachers held assumptions about Black students as being more prone to misbehavior, which could then lead to them perceiving greater misbehavior than was actually present. Teachers might also make more negative assumptions about why Black students were misbehaving even when they engaged in the same misbehavior as White students.”
Was there a way, she wondered, to do a large-scale study to examine whether in fact there was racial bias in teachers’ reporting their students’ behavior problems?
Owens decided that the teachers would have to observe the exact same student behavior in the exact same setting. The only significant difference would be the students’ race. At first she toyed with the idea of planting actors in classrooms, but she decided there were too many variables that would affect the teachers’ perceptions. So she decided instead to create a series of videos.
She began by interviewing teachers about the most frequently occurring classroom disruptions by students of all backgrounds and ethnicities. She selected three vignettes to film: a student texting during a test and continuing to text even after being asked to put the phone away, a student coming in late and slamming the classroom door, and a student throwing objects like pencils or paper airplanes.
Then she hired a video director and cinematographer and held auditions. Her goal was to assemble a group of teenage boys who were similar in height, weight, and build, with the same amount of piercings and facial hair, except that some were Black, some were White, and some were Latino. “I wanted people to be as comparable as possible,” she says, “except for their racial and ethnic background.” She ended up with 14 actors to play the troublemakers, and another ensemble to play the other students in the class.
There are assumptions people make about the reasons for why misbehavior occurs. For White students, the narrative is often that they’re just having a bad day. For Black students, it’s assumed that they don’t care about school and that’s why they’re acting out.
Each video was about 45 seconds long and featured the same disruptive behavior and the same reaction from the actors playing the teacher and other students. The only difference was the race of the lead actor. Owens did a series of pre-tests with a small group of teachers who knew that the study was about race and bias to make sure the behavior was indeed exactly the same across the board.
Finally, Owens showed the videos to 1,339 teachers from 295 high schools and middle schools across the country. The teachers were randomly assigned the race of the student troublemaker in the video. Instead of asking the teachers to rate the behavior on a scale, Owens asked them to describe, in a handful of sentences, what they saw. When Owens and her team got the survey responses back, they coded the language the teachers had used as blaming, empathetic, or neutral.
“We found racial bias on the part of the teachers in the sample,” Owens reports. “Black boys were perceived as being more blameworthy than their White counterparts for the exact same misbehavior in school.”
Latino boys were also perceived as being more blameworthy than White boys, but there wasn’t a significant statistical difference as there was with the Black students. This, Owens says, is consistent with prior research, which has turned up mixed results about whether there’s bias against Latinx students.
Some teachers of color were tougher on students of the same race; others were more lenient. “There may be higher expectations on the part of teachers of color towards students of the same minority group,” Owens explains, “and higher expectations to succeed.”
Owens plans to continue the study using younger students. But she believes her findings reflect biases that affect people of all ages, whether they’re still in school or firmly established in real life.
“There are assumptions people make about the reasons for why misbehavior occurs,” she says. “For White students, the narrative is often that they’re just having a bad day. For Black students, it’s assumed—often incorrectly—that they don’t care about school and that’s why they’re acting out. There’s a difference in meaning-making: even though what you see is the same, the meanings associated with it are very different. And it’s tremendously consequential.”
These consequences can have a rippling effect on these students’ futures: lower grades and rates of college attendance, higher rates of dropout and interaction with the criminal justice system.
“It really suggests that bias is a process that is playing out in this context,” Owens says, “and there are significant effects for downstream outcomes.”