COVID-19 on campus: a New York Times survey and a data void

How is COVID-19 moving across campuses this fall?

We have some data on individual campuses, thanks to the ones that publish coronavirus dashboards. What isn’t so clear is how American higher education as a whole is hosting the pandemic.

A New York Times page/article/tracker gives us some ways of approaching the problem.

A team of around 33 people assembled local data from “more than 1,500 American colleges and universities” into a national account. They didn’t just look at the present epidemic, but included cases going back to early 2020: “over the course of the pandemic, not necessarily where it is prevalent now.”

Top line results: to date there have been “at least 26,000 cases and 64 deaths since the pandemic began.”

(The University of Alabama has protested the Times’ data. They do have a point, and the article responded with an asterisk.)

So, first conclusion: the pandemic has hit higher ed unevenly. Taken together, the virus has certainly found a home in some campuses. I would love to learn more: connections between campus infections and the locality; changes over the course of the past few months; differences between students, faculty, and staff infections; the roles of residence halls and Greek life, to name a few.

Second, this New York Times article/page/tracker is a provisional report at best. Data are tricky to get, and the presented results are almost certainly a substantial undercount. For example,

The Times has counted more than 20,000 additional cases at colleges since late July. Many of those are new infections from this month, but others may have emerged earlier in the pandemic.

As the authors admit, “[w]hile this is believed to be the most comprehensive survey available, it is also an undercount.”

The page/article/tracker can’t determine how many live infections there are on campuses right now.

Moreover, “[s]ome universities just started reporting data, and The Times contacted others for the first time in August.” On top of that, they are only working from about one third of American higher ed, drawing on only “more than 1,500 American colleges and universities.” And “at least 600 others ignored inquiries or refused to answer questions.”

Additionally, research into death statistics compared over years suggests that coronavirus deaths are underestimated. The CDC gives varying results on this score. One peer-reviewed article estimates that “the number of COVID-19 deaths reported in the first weeks of the pandemic captured only two-thirds of excess deaths in the US.” The Times’ number of 64 campus deaths could really point to something closer to 90. If the same undercount problem occurs for infections, that total could stand as high as 40,000. And if the Times only capture one third of US higher ed, both totals could be up to three times higher: 270 and 120,000.

As far as I can tell, flawed and limited as the article/page/thing is, it is the best information we have right now on COVID-19 in American colleges and universities as a sector. The CDC doesn’t publish anything like it — in fact, their higher ed page was last updated one month ago. The Department of Education doesn’t seem to publish any such data, although I have a query in.

Unless there are other projects out there, we’re dealing with a serious data void in understanding American higher education’s encounter with COVID-19. Readers, have you seen anything?

(cross-posted to my blog)