Context
See Andrew's comment.
Further question to resolve: when should we use the 2019 population estimates? E.g. maybe prop signals in early 2020? What would the cutoff be?
Likely depends on the technical definition of the estimate. I.e. are these estimates end of year estimates?