That’s the conclusion of this new article by Bernie Grofman and Jonathan Cervas. From the abstract:

It is well understood that even small differences in population can have a disproportionate impact on representation in the U.S. House of Representatives after a decennial census because of the peculiarities of rounding rules that require integer allocations. While the COVID-19 pandemic can be held responsible for accelerating the trend toward the increased use of mail-in balloting, and it affected the ability of the census to collect in-person information, here we call attention to an unanticipated effect of the pandemic on the electoral process that, as far as we are aware, has never previously been identified. By rerunning the apportionment numbers for all states under the assumption that deaths from COVID-19 prior to the start of the Census had not occurred, we show that New York’s congressional delegation would not have lost a seat. New York was the only state whose House seat allocation was affected by disproportionate COVID-19 deaths.