[As usual for these map-based posts, the bigger your screen, the better.]<\/p>\n\n\n\n
Today, let us consider the single-family detached home (SFDH). Specifically, neighborhoods that are made up of them.<\/p>\n\n\n\n
I’m going to use the same two sets of data that I’ve used in previous posts: the US Census’ American Community Survey<\/u><\/a> (ACS), and the Microsoft building database<\/u><\/a>. The Census provides housing data at the level of census tracts. A census tract is a small, usually permanent geographical division. Typically their population ranges between 1,200 and 8,000, with an optimum size of 4,000. There are about 73,000 census tracts in the United States.<\/p>\n\n\n\n I want to focus on census tracts that are primarily SFDHs. To identify these tracts, I’ve set a somewhat arbitrary cutoff: at least 90% of the homes must be SFDHs. Only tracts that meet this standard are included.<\/p>\n\n\n\n The next step is to identify whether a tract is a “neighborhood”. This is a somewhat vague term, but the reason for doing this is to eliminate certain extremely rural census tracts that are so spread out there is little connection between the homes. For example, this tract in Maine<\/u><\/a> consists primarily of SFDHs, but it covers 2800 square miles, so it can hardly be called a ‘neighborhood’. To filter out such tracts, I’ve defined a standard based on the MS building database: buildings must cover at least 2% of the land. This eliminates very rural tracts that in no way could be considered a neighborhood, like the aforementioned tract in Maine. The 2% standard means that the homes are reasonably close to one another and the area is somewhat like a neighborhood. Typical suburban areas might have 10-15% building coverage; a tract with 2% coverage is pretty spread out but still has a neighborhood feel, or at least is a collection of similar neighborhoods<\/u><\/a>.<\/p>\n\n\n\n Applying these two filters results in about 6600 census tracts (out of an original 73,000 across the country). Let’s take a look at some of the characteristics of these.<\/p>\n\n\n\n First, the average age of homes. Note that this calculation is very rough; the census data provides housing age data in 10-year ranges. Also, the data ranges only go back to ‘built before 1940’, so all homes built in 1939 or earlier are included in the same bucket. So there may be some neighborhoods with extremely old homes that don’t get credited accurately, thus affecting the average. So be it.<\/p>\n\n\n\n Given that, the oldest SFDH neighborhood is this one in Wichita, where most homes date to the 1920s:<\/p>\n\n\n