Matching students with teachers based on race may improve pupils’ academic achievement, according to a new study released by the National Bureau of Economic Research.

The paper, titled “The Market for Teacher Quality,” uses statistical and economic analysis to investigate differences in teacher quality and what accounts for them. The bureau, based in Cambridge, Mass., is a nonprofit, nonpartisan economic research organization.

It won notoriety recently as the place where Harvard University’s president, Lawrence Summers, made now-infamous remarks on women in the sciences. Elements of the race study likewise touch on politically sensitive subjects, some education researchers said yesterday.

The study, which was first made available last Friday, was written by a Hoover Institution fellow, Eric Hanushek; a professor at Amherst College, Steven Rivkin, and the University of Texas at Dallas’s Daniel O’Brien and John Kain.

Evaluating the caliber of teachers based on student achievement in grades five through eight in a large urban school district in Texas, the researchers found that while “good teachers seem to be good teachers for everyone, regardless of whether they’ve got good or bad kids,” pairing students with teachers of the same race seemed to improve student performance, Mr. Hanushek said.

For example, the researchers concluded that on average black students with black teachers “would be boosted 2-4 percentile points,” Mr. Hanushek said in an e-mail.

“What we concluded, holding constant the overall quality of the teachers, was that it did make a difference that black kids have a black teacher, and white kids have a white teacher,” he said in a telephone in terview yesterday.

“What policy should come from that,” he added, “is not obvious.”

Despite the measured advantages of pairing students with teachers based on race, Mr. Hanushek said, “We do not think this should take us back to more segregation.”

The benefits of integration for African-American students, he said, outweighed the possible benefits of grouping students based on race to match them with teachers of their own ethnicity. “We found systematically that if there were higher concentrations of black kids in classrooms, that tended to hurt the black kids,” he said, adding that white students’ achievement did not appear to be affected by integration.

The usefulness of policy proposals at this point based on the findings was “pretty speculative,” Mr. Hanushek said, adding that the subject was one “we’re just starting to think about and explore.”

One possibility, he said, was assigning new African-American teachers to African-American students in their first teaching jobs.

First-time teachers, he said, “tend systematically to be assigned to schools with more disadvantaged and minority kids, and that’s a problem because rookie teachers are not as good in their first year as they will be later on.” He said: “It takes a while to learn what you’re doing in the classroom.”

To mitigate some of the disadvantage that minority students suffer by being taught by a stream of rookie instructors, “If you’re assigning a bunch of new teachers to a school, you might make sure black kids got black teachers,” he said.

Part of why more isn’t known about the benefits of matching students with teachers based on race is a tendency to avoid studies that link student performance to teachers, and to avoid the “very touchy issue” of race in the classroom, Mr. Hanushek said. “There’s a lot of resistance to accountability,” the researcher said, adding that race has, “for all the obvious reasons, been a matter of somewhat divisive discussions since 1954.”

A Manhattan Institute fellow who directs its Education Research Office, Jay Greene, agreed that race is sometimes a hindrance for researchers exploring certain aspects of student achievement.

Mr. Greene said that if the findings of the new study are true, “I’m not sure there’s any policy you would adopt … because either the policy would be politically or socially undesirable, or it would benefit some students at the expense of others in ways hard to predict.”

“It’s certainly possible that there are some questions that are under-examined because no one thinks you can do anything with the results,” Mr. Greene said. Both he and Mr. Hanushek said other research had demonstrated benefits of race-based student-teacher matching.

The president of the New York-based Center for Educational Innovation, Seymour Fliegel, said experience suggested the opposite. He said that the concept of pairing students with teachers of the same race was “nothing new,” and that “the practical experience of following through on this policy hasn’t really played out” in New York.

“I can take you to many schools in New York City that are 99% minority with loads of minority teachers, and not much has changed,” he said.

Mr. Fliegel, a widely praised author who spent many years in the city’s school system as a teacher and administrator, also expressed concern that the study might fuel arguments for segregation.

He said that according to other studies, the most important factor determining teacher success was a teacher’s own academic record, and he said hiring and placement should be based on “who the person is, are they smart themselves, do they have a positive experience in their own studies?” That sort of consideration, he said, was “more important than race.”

An African-American member of the New York City Council, Charles Barron of Brooklyn, disagreed. He said of the bureau study’s findings: “We’ve been saying that all along.”

Speaking of the New York City schools, Mr. Barron said: “We already have segregation. The question is the best quality education for children. If 85% of children are people of color, then 85% of the administrators should be people of color.”

Mr. Barron, who is chairman of the higher-education committee in the council, said currently New York “has just the opposite,” with whites representing more than 70% of teachers, and like percentages of principals and of top aides to the schools chancellor, Joel Klein.

Mr. Barron said the administrative structure of the city’s schools should be “representative of changing demographics,” adding: “People of color are the new majority. We should have a majority of power positions and teaching positions, and this study shows that’s something that should happen.”

The president of the United Federation of Teachers, Randi Weingarten, disputed whether the study showed anything conclusive at all.

As The New York Sun said yesterday in an editorial, “The Market for Teacher Quality” rebutted the claim often made by teachers unions that good teachers leave cities for higher pay in the suburbs.

“Looking at how the data was so flawed on other conclusions,” the UFT president said of the study, she would have to “look at the data really carefully” on the racial questions before taking those findings seriously. In the meantime, however, Ms. Weingarten said that while she has supported single-sex schooling, she is “totally and completely opposed” to any notion of segregation.

“Diversity and integration is the way in which societies grow. When you have isolation and segregation you end up having intolerance, disrespect, and incivility.” Even if separating students to match them with teachers of the same ethnicity improved test scores, Ms. Weingarten said, segregation would not be worth it.

“Our goal here as educators is to help increase academic achievement and to help kids become productive members of society,” she said. “Consequently, if you do one good at the expense of the second, then it would be terrible.”

The New York City Department of Education declined to comment for this article.