Using real world experiments to study mechanisms of inequality in the US and Latin America
Felipe Dias grew up in a working-class neighborhood in São Paulo, Brazil's largest city, and his parents worked multiple jobs to send him to a private school. He couldn't help but notice the contrast in social status between his wealthy classmates and kids like him who didn't come from money.
He didn't know anything about sociology at the time, but these days he thinks that experience might have planted a seed of curiosity about social issues and inequalities. Now he's an assistant professor of sociology in the School of Arts and Sciences—and the new director of the Latin American Studies program—and focuses his research on social stratification and inequality in the U.S. and Latin America.
Some of that research has touched on immigration, which he knows firsthand. As a teenager, he was deeply into basketball and played for his school in São Paulo. A classmate spent a semester as an exchange student in New Jersey, and suggested to Dias that he could play basketball in the U.S.
Dias ended up sending a videotape of himself playing basketball to a Newark, New Jersey, high school basketball coach, who invited him to come enroll and play for his team. He did, and it was a life-changing event. As high school ended, he decided to stay in the U.S. He tried for a college basketball scholarship, but didn't quite have the chops, so he sold vacuum cleaners for a year to make money before heading out to California and eventually UCLA.
By then, academics was his strong suit, and he was on a pre-med track. But he happened to take a sociology class, and then another, and soon switched his major. An undergraduate summer research program in sociology at UC Berkeley cemented his interest, and after graduation he started a Ph.D. program in sociology at Berkeley, focused on race and inequality in Brazil.
As a doctoral student, he focused his research on race in Brazilian labor markets. He decided to take what was at the time a fairly new approach in sociology, focusing on field experiments. It became his hallmark: testing theories in real world settings.
He set up an experiment with fictitious people applying for jobs. While similar field experiments had been done in the U.S. and Europe, he had to design it a little differently for Latin America, because the regions have different histories.
In colonial times, more than 4 million enslaved people were forcibly taken from Africa to toil in Brazil, 10 times as many as were taken to the U.S. But while the U.S. "continued to adopt more restrictive policies around race, over time Brazil adopted a sort of policy of incorporation, integrating Brazil," says Dias.
In the U.S., he says, "racial categories are fairly fixed—you're either white, Black, or Hispanic, for example. But there's also this issue of skin color discrimination, or colorism. Lighter-skinned Blacks tend to do better than darker-skinned Blacks in access to health care, and are treated differently by the criminal justice system and by employers."
In Brazil, Dias says, "colorism is a more prominent feature of race relations, because racial categories are very fluid." While sociological experiments using job applicants in the U.S. rely on using stereotypically white or Black names, and seeing who gets invited for interviews, that wouldn't work in Brazil. "It's more about the skin color gradations there," he says.
So Dias included photos in his fake job applications. His analysis found that darker-skinned job applicants experienced more discrimination than lighter-skinned applicants. What was most interesting was that discrimination was affected by the applicant's perceived class status, Dias says.
That's different from racial discrimination in the U.S. "There's not a lot of intersection of race and class status in the U.S., but in Brazil, I found that racial discrimination is almost mitigated by your perceived class status," he says.
In Brazil, he found that darker-skinned applicants "who are perceived to have a middle-class background were not discriminated against as much as lighter-skinned applicants who were perceived to come from a working-class background," he says.
He also found that skin color discrimination "was largely gendered," he says. Dark-skinned women experienced much higher levels of discrimination compared to lighter-skinned female applicants, regardless of perceived class status—different from what men experienced.
Delving into the inequalities
As a postdoc at Stanford, Dias pivoted to do more U.S.-based research, looking at how organizational features like non-discrimination policies at institutions and organizations might either help reduce or exacerbate racial discrimination. He and his colleagues found and rated diversity and inclusion policies of companies and organizations, to which they then sent fictitious resumes with stereotypically white and Black names.
They found that firms and organizations that publicly listed policies affirming the value of diversity were more inclusive—the policies and statements weren't just for show. "They tend to discriminate less," Dias says.
Now Dias has turned his research interests in the direction of immigration, specifically examining discrimination based on national origin. "Does it matter if you're a foreign-born worker or a native-born worker? Do you get treated the same way, or are there differences?" he asks.
He is working on a large study, again using field experiments. Early findings suggest that there is indeed discrimination over nativity in hiring, though it affects Eastern European immigrants differently than Hispanic immigrants. While employers perceive Eastern European immigrants as lacking English skills, "for Hispanic immigrants, it is a combination of concerns over English skills and perceived cultural foreignness," says Dias.
During the height of the pandemic, he had to postpone some of his immigration-focused data collection and turned to examining the pandemic's impact on social inequality.
He used existing government economic data to see whether parents were more deeply affected by the pandemic than non-parents, and whether mothers were more affected than fathers. A second focus was measuring the pandemic's impact on social attitudes toward Chinese Americans in the U.S. He is hoping to publish that research soon.
In the field experiments he leads, methodology is always important—in the case of the nativity study, for instance, the researchers needed a way to signal whether an applicant was an immigrant or a native-born American without drawing attention to the fictitious applicants.
Dias used three strategies to signal nativity status: names, country of high school graduation, and degree of bilingualism. For example, he says, "U.S.-born Hispanics had an English first name and a Spanish last name; U.S.-born Eastern Europeans had English first and last names; Eastern-European immigrants had a Ukrainian or Polish first name and last name; Hispanic immigrants had Spanish first name and Spanish last name."
Dias has established his expertise in sociology methodology, and regularly teaches quantitative research methods and research design and interpretation courses.
It's all in service of understanding more deeply the nature of social relations at work in the U.S. and in his native Brazil, two places he has seen people struggle with inequalities in different ways. "I think my background informs the kinds of questions that I ask," he says.
Provided by Tufts University