African-American students in schools with high interracial contact may feel greater social pressure to sacrifice their grades in order to maintain popularity than students in predominantly African-American schools, according to a paper published last month by Harvard economist Roland G. Fryer Jr. and graduate student Paul Torelli.

In their study, the researchers focus on the theory of “acting white”—generally understood as the situation in which black students face ridicule from their peers for engaging in behaviors allegedly considered to be characteristic of whites, such as earning good grades, raising their hand in class, reading books, or taking an interest in the fine arts.

Fryer and Torelli also found that while African Americans are often suspected of suffering the most from “acting white,” Hispanic students often face even more social pressure to underperform in class than do black students.

Over the past two decades, some social scientists have argued that African-American students may deliberately avoid behaviors that could benefit them in the future—such as doing well in school—for fear of being perceived as “acting white.”

That behavior, the theory holds, may be at least partially responsible for the persistent academic achievement gap between whites and certain minority students.

“‘Acting white’ very well may explain the lack of minority students in elite colleges and universities,” Fryer said.

Fryer said that, in one important way, the study undermines conventional wisdom on the phenomenon.

“We thought we knew that ‘acting white’ happened in all-black schools in urban neighborhoods. I think we had the facts entirely backwards,” Fryer, who is a junior fellow in the Harvard Society of Fellows and a faculty research fellow at the National Bureau of Economic Research, said in an interview.

In their paper, “An Empirical Analysis of ‘Acting White,’” Fryer and Torelli find that “acting white” is more pronounced in public schools and schools whose student body is less than 20 percent African American. At the same time, the phenomenon is essentially non-existent among black students in primarily black schools, as well as among black students attending private schools. Fryer noted that other factors—such as under-funding, differences in teaching quality, and disparities in parent-teacher interaction—probably explain the differences in overall performance among the nation’s schools.

Fryer and Torelli found that for white teenagers better grades coincided with greater popularity, while black students who excelled in school were substantially less popular than black peers with lower grades.

That phenomenon was even more pronounced among Hispanic students than among blacks, the authors ascertained.

“A black student with a 4.0 has, on average, 1.5 fewer same-race friends than a white student with a 4.0,” Fryer and Torelli write in their paper. “A Hispanic student with a 4.0 grade point average is the least popular of all Hispanic students, and has 3 fewer friends than a typical white student with a 4.0 grade point average.”

Fryer said that the implications of these findings for public policy are not yet clear.

“We’re one step removed from public policy,” he said. “This paper tells us a good place to look.”

The authors generated an economic model defining the trade-off between popularity and academic achievement. They determined that that relationship is “categorically different” among each of the major racial groups.

“The principal idea is that individuals face a two-audience signaling quandary: signals that beget labor market success are signals that induce peer rejection,” the wrote in the report.

Lee Professor of Economic Claudia Goldin, who studies the economic histories of education, income, inequality, immigration, and technological change, praised the authors’ work.

“I have the highest respect and regard for the work of Fryer and Torelli. It is enormously creative and revealing,” she wrote in an e-mail. “Through the lens of their work we have a much better understanding of these behaviors and the functions they serve.”

Fryer and Torelli used data from the National Longitudinal Study of Adolescent Health, which surveyed more than 90,000 junior-high and high-school students from 175 schools in 80 communities around the country. Students taking the survey were asked to define their own ethnicity, Fryer said. Rather than allow students to define their own social status, the authors created a “spectral popularity index” to measure, for each student, the number of same-race friends within his or her school. They then weighted each student’s popularity by the popularity of each friend.

Torelli was not available for comment.

—Staff writer Daniel J. T. Schuker can be reached at dschuker@fas.harvard.edu.