These comparisons are rarely informative, because they're based on arbitrary municipal lines in the sand, rather than at a steady place on the CBD -> Urban -> I

These comparisons are rarely informative, because they're based on arbitrary municipal lines in the sand, rather than at a steady place on the CBD -> Urban -> Inner Suburban -> Outer Suburban -> Exurban -> Rural spectrum A more informative 1-dimensional 1-variable consideration, though imperfect, might be to take every person in the municipality and ask 'How many other people live within 1km of this person?', and take the median of that distribution.