Normalising "subscribers to Martha Stewart Living" and "consumers of furry pornography" by population would give you per-capita values, showing you places where

Normalising "subscribers to Martha Stewart Living" and "consumers of furry pornography" by population would give you per-capita values, showing you places where those properties had high incidence per-person (to which you'd have to apply significance tests). I'm pretty sure it is a normalisation joke.