Hometown:
Boston, Massachusetts
Undergraduate Institution:
Hamilton College ’17
Major:
B.A. Economics and Mathematics
Jaemin Woo, a Ph.D. student in Brown University's Department of Economics and a participant in the Watson Institute's Graduate Program in Development (GPD), is trying to answer a question that one might not usually associate with economics: "How do the songs that people listen to as adolescents affect their preferences, beliefs and values later in life?"
Music's impact from adolescence to adulthood
According to Woo, "People tend to go back to the music that they listened to when they were in their formative years." This is a widely acknowledged phenomenon and studies by psychologists and neuroscientists have confirmed that the music we listen to during adolescence tends to hold a powerful influence on our emotions, and the connection doesn't weaken over time. Given its potent neurological influence, it stands to reason that the music we listen to in our formative years might affect our attitudes, values and beliefs as adults. In his Ph.D. research, Woo is trying to quantify that connection.
"Because I'm an economist," Woo said, "I'm inherently driven to try to find a causal relationship between what we listen to and how that affects our beliefs and preferences." However, he notes it is challenging to disentangle cause from effect when studying musical listening patterns.
For example, Woo says there might be a connection between people who like reggae music and people who want to protect the environment. "It could be that people who favor environmental protection are drawn to reggae music," he said. "We don't know if listening to reggae songs leads people to become environmentalists or if they choose to listen to reggae because they already are."
"The only foolproof way to do this," said Woo, "would be to conduct an experiment where you took 100 babies and exposed 50 of them to reggae music and 50 of them to something else, and then compared the two groups after they've grown up." "Of course," he added, "a study like this would never pass IRB [Institutional Review Board] protocols."
Instead, Woo said he is using the statistical frameworks and data that are available to him to try to "tease out" a causal relationship.
Woo is using Chat GPT to evaluate the lyrical content of all songs on the Billboard Hot 100 from 1958 to 2023. "One thing we're doing is rating the level of support for feminism in lyrics," said Woo. "So we are able to quantify the level of support for feminism in all the songs that were popular during a given period."
Woo said that Chat GPT has been highly successful in measuring different nuances in lyrics and musical content which has given him a huge amount of data to work with. "We gather all these data from Spotify's database, and together with AI we can measure musical metrics. Basically, we're able to translate songs into numbers," he said.
Woo then compares these data against data taken from the biannual General Social Survey which keeps a historical record of the attitudes of U.S. residents. "The beauty of this is that we can run natural experiments on thousands of people without having to actually expose them to music," he said.
Using this methodology, Woo said he has found some statistical evidence that suggests a causal correlation between the music that was popular during a person's adolescence and their attitudes later in life. "It does seem to have an effect on your preferences and subjective sense of well-being," he said. "People in other fields like musicology and sociology have talked about this and we're trying to confirm those claims using this large database and methods that are empirically based."