[As usual for these map-based posts, the bigger your screen, the better.]
I’ll continue the analysis of building and census data that I started in my previous post. Today I introduce a new metric, which determines how close together buildings are in a given census tract. For each building in a tract, we measure how many other buildings are within 100 meters (312 feet)1Details: define the center of each building (the centroid of the polygon). Draw a circle centered on that point, with a radius of 100 meters. Calculate how many other building centers are in that circle.. Here’s an example, for our previous home in San Diego:
Counting up the red buildings, we find 32 homes within 100 meters of this one. Each building in a census tract has a value; if you take the average of these values for a tract, that gives you a good indicator of how close together the buildings are. I’ll call this metric Average Nearby Buildings (ANB). Turns out the ANB for this neighborhood is around 27.
The nice thing about ANB is that it doesn’t matter how big the census tract is. There can be a lot of empty space, but if the buildings are all packed in close together (or spread out), it will show up in this number.
Let’s start with the tract with the highest ANB: a seaside community in Manhattan Beach, California:
A typical building in this tract has 106 nearby neighbors. This neighborhood basically consists of 500 duplexes, each with a footprint of about 1000 square feet. The buildings are very close together, without front yards, and most without backyards either. I guess when you’re that close to the beach, you don’t need much else:
Here is a summary of the top 100 census tracts, ranked by Average Nearby Buildings. The list consists mostly of Chicago neighborhoods, very much like the ones I looked at in my last post. Chicago dominates this list, because it has scores of neighborhoods that look alike: small houses, close together, with detached garages (which adds to the building count). Here are the top 100:
Rank | Location | ANB |
1 | Manhattan Beach, CA | 106 |
2-25 | A bunch of Chicago neighborhoods. Here is a typical one. | 105-97 |
26 | Long Beach, Long Island, New York, a community similar to Manhattan Beach | 96 |
27-54 | More Chicago neighborhoods | 96-89 |
55 | A mobile home park in Mesa, Arizona | 89 |
56-76 | More Chicago neighborhoods | 89-87 |
77 | Another mobile home park in Mesa, Arizona | 87 |
78-100 | Still more Chicago neighborhoods | 87-85 |
The Chicago tracts are residential, but you can’t really define them as classic “single family detached home” neighborhoods. That’s because many of the homes are multi-unit. They don’t look it from the outside, but they are divided into two or sometimes 3-4 units.
What about on the other end of the spectrum? That would be census tracts whose buildings are spread out, not near other buildings. A few tracts make this list only because they have so few buildings. For example, this high-rise condo complex in Alexandria, Virginia, has just five buildings, and its ANB is 0.4.
Let’s ignore tracts with a tiny number of buildings. Filtering those out, we find this census tract in rural Virginia. Just 80 miles from the condos in Alexandria, and it couldn’t be more different:
There are 2252 buildings in this tract, and the typical one has just one neighbor within 100 meters (AVB is 1.2). Here is the ‘hot spot’ of the area, with seven buildings all within a 100 meter radius. Everywhere else is less crowded
Average Nearby Buildings provides an overall value for a census tract. But each building has its own value. Within each tract one building has to have the highest value (like the one above, with 7 nearby buildings). In fact, we can look at all of the 125 million buildings in the Microsoft database to determine which one has the most nearby buildings. Drum roll please…
It’s on the Jersey Shore, and it has 180 neighbors within 100 meters:
One of many many beach bungalows jammed in close together, even tighter than Manhattan Beach:
Congratulations, 21 East Bay Way, Lavallette, New Jersey!
Next time, I’ll focus on neighborhoods that are made up of that staple of American housing, the single family detached home.
Interesting that beach areas, Chicago, and mobile home parks all have high ANBs, yet probably have people with very different life-styles and incomes.