[craigs]

Originally Posted by ChiSoxRox

Thanks for the kind words. I did really want to see how the big metros shook out, but this thread's interest helped with motivation.

Hartford is one of the more shrunken Northeast cities, and was iirc the second largest in BosWash to still shrink in 2020, after Baltimore. Hartford proper has several <5k tracts interspersed between the 15k+ needed to really pull up the WPD, while its suburbs quickly drop to 2k ppsm or less. New Britain and Storrs have one tract each over 10k, while whole suburban towns don't touch 5k. Hartford proper is also only 60% of Providence proper's size -- a smaller core in the first place.

Meanwhile, Providence's MSA also inclues Bristol County, MA -- New Bedford and Fall River, meaning Providence MSA has three dense cores to work with. Providence proper -- growing 7% -- is a quilt of 10k and 20k tracts with the edges staying above 5k, while New Bedford and Fall River also have impressive density in their cores. Then, just eyeballing the Census map, Rhode Island suburbia looks more compact with a quicker gradient to rural fields than Connecticut cul-de-sacs.



SF + SJ: 12,025 ppsm
LA + IE: 10,255 ppsm

Combining MSAs isn't tricky: (WPD_1 * Pop_1 + WPD_2 * Pop_2)/(Pop_1 + Pop_2)

Finally: What bastion of urbanity is the biggest MSA to have a WPD that doesn't even clear 1,000? Turns out it's a place that likely sees a few New Yorkers:

Myrtle Beach, SC - 958.2 ppsm.

Thanks for all your hard work. Do you think you might eventually get around to doing weighted density for CSAs?