If you are black and looking for a place to rent, you may be out of luck the moment you leave a message on the leasing agent’s telephone.

That’s what a team of undergraduates discovered when they served as testers for a research project headed by Dorothy S. Thomas Professor of Sociology Douglas Massey and postdoctoral fellow Garvey Lundy. The researchers discovered significant race, class and gender discrimination in the Philadelphia housing market based solely on the speech patterns of the would-be renters.

Massey and Lundy published their research results in the March issue of Urban Affairs Review. Expanding on an earlier study that documented similar discrimination in the San Francisco Bay area, their team set out to measure discrimination based on calls to Philadelphia-area landlords and rental agents.

Students in Massey’s course, Studying Racial Discrimination, made 474 phone calls to 79 agents who advertised in the Sunday Philadelphia Inquirer real-estate section and two local rental guides. The students followed a standard script with identical life histories, incomes and rent requirements. But they assumed three different speech patterns, two middle class — white middle-class English and black accented English — and the lower-class dialect known as Black English Vernacular.

While 76 percent of the males who spoke white middle-class English were ultimately offered an apartment to inspect, only 63 percent of the men who spoke black accented English were. The women fared worse — 60 percent of those who spoke white middle-class English and 57 percent of those who spoke black accented English were offered apartments. And of those who spoke Black English Vernacular, only 44 percent of the men and 38 percent of the women were offered apartments.

Many of the students, Massey said, were surprised by what they discovered. “The white students were more surprised than the black students, the women were more surprised than the men, and the white women were particularly surprised at the extent of gender discrimination.

“The blacks were also surprised that the discrimination was more blatant than they thought it was.”

This, Massey said, shows how technology — in this case, voice mail and answering machines — can make it easier to discriminate. “You can discriminate without ever speaking to or seeing someone simply by not returning a phone call. It’s harder to discriminate when someone is sitting in front of you.”

While scholarly studies cannot be used as law-enforcement tools, Massey hopes fair-housing advocates will use his methodology to conduct tests of their own.

“[This study] shows the persistence of illegal behavior and the need for enforcement of fair-housing laws in the United States,” he said.