Gerard Torrats-Espinosa’s recent research leverages machine learning to examine the variety of factors that shape the spread of COVID-19, including the role of residential racial segregation.

Gerard Torrats-Espinosa
Gerard Torrats-Espinosa

For Gerard Torrats-Espinosa, academic research in urban sociology and stratification is rooted in his experience as a firefighter in Barcelona, Spain. Many Spanish firefighters, he explains, are recruited from schools of engineering and architecture with the idea that their structural knowledge will help preserve and reinforce buildings. Torrats-Espinosa describes the nearly five years he spent as a firefighter as one of his best life experiences. Firefighting allowed him to apply his own engineering background towards public service, and a number of questions arose around the patterns of intervention he witnessed in more disadvantaged neighborhoods.

This interest in the relationship between people and place and his search for answers at the intersection of urban planning and social policy led Torrats-Espinosa to the U.S. He earned a master’s degree in public policy from the Harvard Kennedy School of Government in 2014, and a doctorate in sociology from NYU in 2019. His doctoral dissertation looked at the geography of educational opportunity in the U.S., with a nationwide exploration of how neighborhood violence affected the academic achievement of local children—specifically, on how reductions in violent crime led to improvements in test scores of racial and ethnic minorities. “Understanding inequalities in academic achievement requires evidence on what happens outside, as well as inside, schools,” he explains. 

Support from the Data Science Institute (DSI) enabled Torrats-Espinosa to publish work from his dissertation as a postdoctoral researcher, and he joined Columbia’s faculty as an assistant professor of sociology in Fall 2020. He explores the intersection of space, place, and social outcomes. He examines how social capital and social organization emerge and evolve in spatial contexts, with a focus on how geography—places and spaces—affect educational and economic opportunity in the U.S. and elsewhere.

In the midst of the coronavirus pandemic, Torrats-Espinosa turned his eye towards the intersection of people, place, and health to research the effect of racial segregation on COVID-19 infection and mortality rates in the U.S. In a recent paper published in the Proceedings of the National Academy of Sciences titled “Using Machine Learning to Estimate the Effect of Racial Segregation on COVID-19 Mortality in the United States,” he leverages machine learning to examine the variety of factors that shape the spread of COVID-19, including the role of residential racial segregation. Like his previous work on community crime rates and educational outcomes, this paper aims to unpack the concept of “place”, and examines the way that “place” intersects with other factors.

Torrats-Espinosa, who is also a DSI member, assembled a data set that includes 50 county-level factors that are potentially related to COVID-19 infection and mortality rates, such as demographics, density, social capital, health risk factors, capacity of the health care system, air pollution, employment in essential businesses, and political views. He used machine learning to guide the selection of the most important controls and highlight the most statistically significant variables in a manner that epitomizes his approach to research: “Transparent, honest, and statistically principled.”

The findings indicate that more segregated counties had higher mortality and infection rates overall and larger mortality rates among Blacks relative to whites. This work attests to the importance of including features of the built environment in calculations and mathematical models that forecast the spread of infectious diseases. “It’s very easy to blame people for outcomes, even though we know how factors such as the place where you grew up and access to resources matter,” Torrats-Espinosa said. “This work highlights the impact of place and the built environment. I hope that these insights can inform and improve public policy.“

Torrats-Espinosa expressed his gratitude to DSI for enabling him to focus on publishing his dissertation research, and for his exposure to the range of fields and ways to engage with data across DSI and Columbia. “I’m inspired by the variety of ways that the academic community is embracing computational methods and data science, with DSI and Columbia at the forefront,” he concluded.

— Karina Alexanyan, Ph.D.