Robert Rosenthal and Lenore Jacobson famously experimentally manipulated teachers’ beliefs of student ability by providing false information regarding students’ performance on a nonexistent test. The researchers found that students who were falsely identified as “growth spurters” experienced significantly greater school-year learning gains than their classmates.
Today, anecdotes abound about the importance of teachers believing in students’ ability to succeed. For example, a third grader in Pittsburgh noted “My favorite teacher is Mrs. Patty. She’s my favorite because she believes in me even when I’m doing badly . . . There was a time when I couldn’t do cursive and Mrs. Patty helped me do it. It wasn’t hard because she believed in me.”
When teachers believe in students, they tend to succeed. That’s what makes it so troubling that research shows teachers’ educational expectations for racial minority students are significantly lower than they are for white students. For example, tenth grade teachers are twice as likely to expect black students to complete no more than a high school diploma as they are to have similarly low expectations for white students. However, it is unclear whether these “racial educational expectation gaps” are evidence of systematic biases in teachers’ expectations or simply reflect accurate predictions (perhaps due to racial differences in access to school quality or early childhood investments)..
Three researchers address this question in a new Upjohn Institute working paper. Seth Gershenson, with Stephen B. Holt and Nicholas W. Papageorge, identify systematic biases in teachers’ expectations using data from a nationally representative survey of U.S. tenth graders. The survey asked two teachers per student how much education they expected the student to complete. The authors' research design was sparked by the intuition that if two teachers simultaneously have different educational expectations for the same student, at least one of the teachers’ expectations is wrong. If these differences are systematically related to the demographic match between students and teachers, this suggests that the differences are not random, but are based on systematic bias.
For black students, this is exactly what they find. For example, non-black teachers are about 30 percent less likely to expect black students to complete a four-year college degree than are black teachers. A limitation of this study is that the authors cannot determine whether the black teachers surveyed were too optimistic, the non-black teachers were too pessimistic, or some combination of the two. Nonetheless, evidence of systematic bias in teachers’ educational expectations is concerning, as it likely contributes to persistent socio-demographic gaps in a variety of educational and socioeconomic outcomes.
Importantly, the results are not intended to denigrate teachers. Biases in expectations are generally unintentional and are an artifact of how humans categorize complex information. Instead, the findings highlight the need to better understand how teachers form expectations, what interventions minimize biases in teacher expectations, and how biased expectations affect student success.
Seth Gershenson is an assistant professor in the School of Public Affairs at American University. Stephen B. Holt is a PhD student in the Department of Public Administration and Policy at American University. Nicholas W. Papageorge is an assistant professor of economics at Johns Hopkins University.
 Source: Authors’ tabulations of the Educational Longitudinal Study of 2002 (ELS).
 See, for example, Cosmides, Leda, John Tooby, and Robert Kurzban. 2003. “Perceptions of Race.” Trends in Cognitive Sciences 7(4):173–179.