The other day I was experimenting with the Microsoft Buildings dataset, trying to better understand where there are most buildings. I knew suburban areas had more buildings then urban areas, with sprawling single family homes, a trend you can see around most of the major upstate cities. Manhattan, has a massive under-count of buildings, by almost a factor of 10, due to the way the Microsoft Building algorithm sometimes groups together buildings in urban areas.

How I made this map:

1. Took the Microsoft Building Footprint data, converted to Centroids.
2. Ran Count Points in Polygon, against Census TIGER/Line for County Subdivisions

   https://github.com/Microsoft/USBuildingFootprints