New Publication | May 14, 2019

What musical taste reveals about cultural identity

© aviindy / photocase.com

Using Facebook data on musical preferences, MPIDR researchers have shown in a new study that immigrants in the US adapt to the culture of their new homeland in very diverse ways. Among their findings is the discovery that – against all odds – first-generation immigrants have become highly assimilated.

In a study published today in the Proceedings of the World Wide Web conference 2019, MPIDR Director Emilio Zagheni and MPIDR scientist Tim Riffe investigated how Mexican migrants who entered the US after 1965 have adapted to the culture of their new homeland. To assess the migrants’ cultural assimilation levels, the researchers used data on musical interests from the social network Facebook.

"The factor of musical taste is very well-suited for studying cultural assimilation," Zagheni said. "Music is a marker of cultural identity and social status. It is a key symbolic resource that individuals draw on to negotiate and express their social position." The study was based on the assumption that a population of migrants are more assimilated if their interests have become more similar to the interests of the population in the destination country, and less similar to the interests of the population in the country of origin. "In our study, the term assimilation is used in a technical sense to refer to the reduction of distance between two groups," Zagheni explained.  

To gather Facebook data, the researchers used a special tool developed for the advertising industry. To enable advertisers to reach their target audience, Facebook provides them with aggregate-level data on some demographic parameters and interests of its users. Advertisers can then, for example, stipulate that their ads should target specific migrant groups or specific age groups, or should be visible to users in certain regions only. Advertisers also have the option of selecting who is targeted by their country of origin, their language preferences, and their music preferences. In addition, advertisers are granted access to estimates of the number of people in each subgroup they want to target.

For their investigations, the scientists distinguished between more than 700 different styles of music, and identified the genres that are more prevalent in specific cultures or population groups. For example, Mexicans tend to listen to salsa and mariachi, Anglo-Americans typically listen to rock and pop, and African-Americans often listen to rap. The researchers then made comparisons between groups, such as between first- and second-generation Mexican immigrants and Mexicans living in Mexico, as well as between the two immigrant generations and Anglos and African-Americans.

Their results show a very heterogeneous pattern of assimilation. The analysis revealed, for example, that first-generation Mexican immigrants have assimilated to a significant degree to both Anglos and African-Americans. The findings also indicated that assimilation levels are high among second-generation immigrants, but lower among the sub-group of those immigrants who use Spanish as their primary language. Moreover, the study showed that the level of education has a considerable effect: i.e., as a general rule, the higher an immigrant’s level of education is, the more assimilated he or she is likely to be.

The researchers also looked at regional differences, and found that Mexican immigrants have musical interests that are more similar to the ones of African-Americans than to the ones of Anglo-Americans. A potential explanation for this finding is that immigrant groups are more likely to come into contact with African-Americans, based on their residential opportunities and choices. It is also possible that because immigrants are on average socially disadvantaged, the musical taste of other socially disadvantaged groups resonate more with them.

"Our study shows that assimilation is not a uniform and linear process that applies to all immigrants, even if they have the same country of origin," Zagheni said. The findings also suggest that the so-called "straight-line theory" is not applicable to all migrant groups. The "straight-line theory" is based on research done on the first wave of immigration from Europe to the United States at the beginning of the last century. Several studies have shown that most European immigrant groups became largely indistinguishable from Anglo-Americans after a few generations. It was assumed that this process is a fundamental characteristic of assimilation that is applicable to all immigrant groups. However, this theory did not hold when it was applied to immigrant groups from Latin America and Asia who came to the United States in the 1960s or later. While wealthy migrants quickly adopted Anglo-American habits, immigrants from lower socioeconomic groups tended to take on the cultural norms of the African-Americans who shared their relatively poor neighborhoods. The current MPIDR study also shows that because assimilation is a very diverse process, a single theory cannot be applied to all immigrants.

In addition to providing new results, this study used new methods. The use of social media data for social science research is still in its infancy. Traditional demographic research relies mainly on "high quality" data, e.g., on census data or other data collected by official agencies, or on data derived from sampling surveys. While the quality of the Facebook data cannot be fully guaranteed, using these data offers completely new insights that classic data collections cannot. For example, in previous assimilation studies, comparisons were only possible between first- and second-generation migrants. "Facebook’s worldwide popularity allows us to directly compare the cultural preferences of Mexican immigrants with those of Mexicans living in Mexico," Zagheni said. "This comparison is important because it challenges the notion that first-generation migrants are a suitable reference population by showing that these immigrants may be more assimilated than is commonly assumed. There was no way to make these comparisons using traditional data sources." He also pointed out another advantage of the Facebook data: "Because the interests are defined by the users, these data provide a bottom-up perspective on cultural interests that is often overlooked in other data sources, such as surveys."

Zagheni also emphasized that the study did not use personally identifiable information, noting that he and his colleagues "only used aggregated data that are routinely produced for advertisers, and that do not allow us to draw any conclusions at the individual level."

The article is the result of a research project that involved not just MPIDR researchers Emilio Zagheni and Tim Riffe, but René Flores from the University of Chicago, Ingmar Weber from the Qatar Computing Research Institute, and Ian Stewart from the Georgia Institute of Technology. Ian Stewart worked on the project when he visited the MPIDR as part of the MPIDR guest program in summer 2018.

More Information