Colleges that reopened for in-person instruction this fall probably contributed more than 3,000 Covid-19 cases a day in their counties that wouldn’t have emerged if they’d remained online, according to a report released on Tuesday.
The report was based on a study conducted by researchers in epidemiology, health economics, and higher education at Davidson College, Indiana University at Bloomington, the University of North Carolina at Greensboro, and the University of Washington. It found that reopening for face-to-face instruction, which prompted far more students to return to campuses, resulted in 1,000 to 5,000 additional cases per day, with the best estimate around 3,200.
As expected, when students moved from counties with high infection rates, the campuses where they arrived experienced more Covid-19 cases, according to the report, which is due to be posted on a health-services preprint server on Tuesday.
The increase in case counts largely stemmed from more people flowing into and around the campuses during the week just before classes started and for two weeks afterward, the researchers said. They followed those movements using students’ cellphone GPS-tracking data. The researchers calculated infection rates in the surrounding counties from July 15 to September 13, before and after students arrived.
The study, which has not yet been peer-reviewed, nonetheless has important policy implications, Christopher R. Marsicano, an assistant professor of the practice of higher education at Davidson, said in an interview on Tuesday.
Colleges shouldn’t send students home this fall, unless it’s the end of the semester, the report recommended. “Once they’re on campus, they should stay on campus,” it said.
Once they’re on campus, they should stay on campus.
Colleges should limit, as much as possible, the movement of students to and from their campuses, the report added. Marsicano acknowledged that, at large universities, that is “extraordinarily hard, if not impossible,” to do. Students who moved from areas with high infection rates to college campuses could become “double spreaders” if colleges whose own infections rose as a result sent them back home, he said.
A number of colleges that have reopened in person this fall have announced plans to continue the semester online after Thanksgiving, when many students are expected to return home.
Plans for the spring semester are still up in the air on many campuses. The researchers said their findings should prompt colleges to look more carefully not only at health conditions in their immediate surroundings, but also in the areas their students are from.
“If you draw a lot of students from Florida or other states with high incidence, it’s maybe worth putting a pause button on in terms of bringing students back,” Marsicano said.
“We’re not wagging a finger at colleges and saying this was a bad idea,” he said of the decision to resume in-person classes. Rising case counts are just another factor that colleges should consider in the cost-benefit analysis of welcoming students back to campuses in the spring versus staying online.
The researchers said they didn’t yet know whether measures colleges had taken to try to mitigate Covid-19 outbreaks, such as the University of Notre Dame’s two-week pause in midsemester, had succeeded, but they hope to determine that.
‘Somewhat Terrifying’
The researchers used a method that looks at differences between two groups and how they change over time. In this case, they examined trends in county and campus case counts before and after campuses reopened. When they saw trends start to diverge between colleges that offered classes in person and those that remained online, and after checking other factors, they concluded that the discrepancy was probably due to the decision to teach face to face.
The researchers had suspected that the case counts would rise when face-to-face classes resumed, but being able to quantify by how much is important, Marsicano said. He called the increases they found surprising and “somewhat terrifying.” Marsicano is also founding director of Davidson’s College Crisis Initiative, which is collaborating with The Chronicle on a regularly updated database of colleges’ reopening models.
Andrew W. Friedson, an associate professor of economics at the University of Colorado at Denver who has studied so-called superspreader events, said it’s too early to do a deep dive into the findings before the study is peer-reviewed. But from what he could see, he said, the researchers are “on the right track.”
“This study uses tried-and-true methods of public-policy analysis,” he wrote in an email to The Chronicle. “Their general approach is appropriate, and I imagine that their findings are in the correct ballpark.”
Josh Michaud, associate director for global health policy at the Henry J. Kaiser Family Foundation, wrote in an email that the study’s overall approach appeared sound.
“Still, multiple factors could contribute to whether these areas had cases increase, and college openings are just one,” he wrote. “It’s possible colleges that reopened were more likely to be located in counties where broader community transmission was higher during the study period for reasons other than the colleges’ being there.” Michaud, who also teaches global health policy at the Johns Hopkins University, added that there might also have been more testing in areas where colleges were reopening — a possibility the authors acknowledged.
“Even so,” he said, “the study does provide some evidence that having students return to college campuses in person did lead to additional spread of coronavirus.”
“Colleges,” the report says, “have done a lot to try and reduce the risk of Covid-19 on campus and in their surrounding communities. Our results in no way say that they have failed or not done enough. Rather, fighting Covid-19 is hard, and more work is needed to understand which methods work best.”
