Tanvi Misra is a staff writer for CityLab covering immigrant communities, housing, economic inequality, and culture. She also authors Navigator, a weekly newsletter for urban explorers (subscribe here). Her work also appears in The Atlantic, NPR, and BBC.
MIT Media Lab’s new interactive “Atlas of Inequality” shows that “segregation is not just about where you live, but what you do.”
When I lived in my old D.C. neighborhood of Mount Pleasant, it was at that particular stage of gentrification where it seemed truly diverse. Taquerias and pupuserias stood right alongside indie theaters and grungy dive bars; the sidewalks were a multicultural mix of young, mostly white professionals and working-class people of color. But if you looked closer, you’d notice what some experts call “micro-level segregation.” People from different economic and racial backgrounds didn’t frequent the same bars, restaurants, and stores. Latinx residents seemed to hang out at Marleny’s, whereas more affluent newcomers would be seen at Marx Café—right next door.
In a new map, MIT Media Lab visualizes that kind of micro-level segregation in the Boston metro region to show that “economic inequality isn't just limited to neighborhoods,” as the researchers write on the website. “It’s part of the places you visit every day.” The map, which the MIT team hopes to expand to the 11 largest U.S. cities, is a part of ongoing research into how individual decisions and opportunities shape real-word urban issues so that “we can act and intervene in human behavior,” said Esteban Moro, the principal investigator at MIT Media Lab and an associate professor at Universidad Carlos III de Madrid.
To create the map, Moro and his colleagues compiled aggregate location data collected by Cuebiq’s Data for Good initiative, harvested from digital devices (like cell phones and tablets) of 150,000 anonymous sources between 2016 and 2017. Based on the median income of the census block group where each anonymous user spent the most time, the researchers assigned them to one of four income brackets. They also obtained a list of 30,000 places—including restaurants, bus stations, museums, offices, and coffee shops—that these users frequented most often. For each place (represented as one dot on the map), they were able to determine its share of visitors from each income category.
Based on that distribution, they placed each place on an inequality index (displayed on the top left corner): The most unequal places (in red) were those where only one type of income group visited in the time period; the most equal (in blue) were those where all four income groups had similar shares—meaning that people of diverse economic backgrounds spent time there at roughly at the same rate.
The resulting map looks a lot like a view of Boston from an airliner on final approach over the city. But the multicolored points of light are actually schools, businesses, and other meeting places.
The resulting “Atlas of Inequality” reveals a taxonomy of places in the city that tend to be more diverse and those that tend to be more economically homogenous. Among the most equal places, Moro and his colleagues found, are museums and airports. Schools, on the other hand, are among the least.
What’s striking, although perhaps not entirely surprising, is that two places can be just meters apart and have a completely different economic profile of visitors. Where we get coffee, where we buy groceries, and where we grab take-out often reflect our choices, which determine the kinds of people we interact with every day. Or, these habits reflect our constraints—and show what places are accessible and welcoming to certain groups of people.
“Right now the way we understand segregation is at the census tract level,” Moro said. “But our decisions that are impacting segregation happen actually at much smaller level—within 25 meters.”
Here’s an example from Boston of two coffee shops (whose names have been anonymized by the researchers to protect the businesses) just across the street from each other, one of which is much more diverse than the other:
The map and accompanying research, of course, have limitations. While Moro and his colleagues made sure that the sample of anonymous users they analyzed was as representative of the general population as it could be, it does nevertheless leave out people at the lower extreme of the income spectrum—people who are homeless, for example—who bear the brunt of segregation. The researchers also acknowledge that the list of places they feature is not comprehensive.
Nevertheless, the atlas offers a very unique and highly detailed close-up of how racial and economic segregation manifests in the city—and how we may have internalized its effects. “Maybe segregation is not just about where you live,” said Moro, “but what you do."
Correction: This article has been updated to reflect that MIT researchers assigned income categories based on median income of the Census block group where users spent most time, not Census blocks.