This week, a team of demography researchers published a paper sharing excess death estimates by county, for the first two years of the COVID-19 pandemic. The team, led by Andrew Stokes at Boston University, has been analyzing excess death data for years in order to understand the true toll of COVID-19 on the U.S.
To measure excess deaths, researchers compare the number of deaths that they’d expect to occur in a given place, over a given timeframe—based on modeling from historical data—to the number of deaths that actually happened. This metric is a helpful one for COVID-19 research, because official COVID-19 deaths are undercounted for a variety of reasons. (To name a few: lack of standards for death certificates, politicization of the pandemic, health equity issues.)
Especially now that official COVID-19 data are becoming less and less reliable, I see excess deaths as a useful avenue for continued reporting on the pandemic. And for any journalists or researchers interested in looking into this issue, Stokes and his team’s work is a great starting point. I collaborated with them for MuckRock’s Uncounted project, using a preprint iteration of the paper published this week.
For more info on this topic, see the Uncounted project and this 2021 post about excess death data from the CDC.