Bias against black students is deeply ingrained and, says an economist looking at national data, a waste of human capital.

Share story

When looking at the same students, teachers of differing backgrounds have vastly different expectations, and new research from Johns Hopkins University suggests that this is no fluke or quirk, but a deeply ingrained problem with profound economic implications.

Nicholas Papageorge, who studies labor and the economy, tracked 8,400 high school sophomores and found that teachers showed a clear pattern of low expectations for their black students.  He and study co-author Seth Gershenson asked two different teachers, who had each taught a particular youth in math or reading, to predict how far that student would get in school.

For white kids, the ratings were similar. But with blacks, particularly boys, there were big differences. White teachers, Papageorge writes, were almost 40 percent less likely than black teachers to believe African-American students would graduate from high school. They were 30 percent less likely to think black students would make it through college.

“When white teachers and black teachers systematically disagree about the exact same student, one of them has to be wrong,” said Papageorge, who previously taught elementary school in New York City.

Education Lab took a few minutes to talk with Papageorge about the meaning of his research.

Q: Why do you think a person’s expectations matter? Why study this?

A: People make investments based on how they think those investments are going to play out. As an educator, if I think a kid doesn’t have a lot of potential I might put my investments elsewhere. So if there are these biases, that’s worrisome because it could be wasting human capital. Unless you are a person who thinks black kids are just not capable — and I’m not — something is really wrong. We are wasting potential. As an economist, it’s what we call ‘inefficient.’

Q: What drew you to this type of research?

A: The original project was trying to understand the impact of teacher expectations. That’s really hard because if I see a kid who does poorly and the teacher says so, it could mean that the teacher is accurate. Or it could mean that the teacher had a low expectation that was communicated to the student, and then the kid says ‘I’m not very smart’ and stops trying.

Q: So how did you get around that?

A: We decided to first see if there was evidence of systematic bias, which we could do by having two teachers — one black, one white — looking at the same student.

Q: But couldn’t family income be a factor in a teacher’s expectations, especially about college?

A: Yes, absolutely. That’s why having two teachers was so important. It’s the fact that we’re looking at the exact same students and having the teachers disagree on such a large scale. There shouldn’t be this magnitude of difference. That’s what did it for me. When we’re looking at thousands of students across the United States and seeing this, well, that’s really big.

Q: Now that you’ve detected this bias, does your research suggest any way of overcoming it?

A: There’s a lot of talk about increasing diversity in the teaching force. I am a little bit skeptical of that. We would need a really large proportion of African-American teachers if we wanted kids to only be taught by educators of their own race. A much cheaper and very overlooked policy is to make people aware of their biases. It can be done. It’s not pretty, but it works. Demonstrating for teachers that they have a lot of implicit biases — just like everyone else in the world — could help mitigate these problems.