Where are the suburbs?

1st Avenue west of Stoney Point Road: still Cedar Rapids, but definitely suburban

[NOTE: This post has been extensively revised in July 2020 due to numerous problems with the original data.]

A few years ago, I referred to the area around Stoney Point Road SW as "in the suburbs." The person I was talking to laughed, because, of course, Stoney Point Road is still within the city limits of Cedar Rapids, and anyhow this isn't like Chicago where they have The Suburbs.

However, there seems to be some analytical value in identifying differences in how and where people live, and it should be obvious to anyone who lives anywhere that we can do a lot better than corporate political boundaries. We can probably do a lot better than census tracts, too, but I'm not going to; at any rate they make a more finely-grained first crack.

City Lab recently reported on a working paper from the Harvard University Joint Center for Housing Studies that group analysis of the urban-suburban divide into three broad approaches:

1. census-convenient: Urban means the principal city of a metropolitan area, as well as any other towns of more than 100,000 population; suburban means all other political entities in the metro. (By this definition, 70 percent of Americans live in suburbs.) (My Illinois readers will note this classifies the entire City of Naperville as "urban." I'll wait here while you finish laughing.)
2. suburbanisms: Urban means places where the proportions of commuting by private car, homeownership, and single-family housing are below the average for the entire metropolitan area, or where any two of the three are below the metro average; suburban means where commuting by private car, homeownership, and single-family housing are more the norm. (By this definition, over 60 percent of Americans live in suburbs.) (That this is lower than definition #1 surprises me, because I live in Cedar Rapids, where much of the political city is actually suburban, but I imagine in more complex metropolitan areas like Chicago many areas outside the political boundaries of the city would also be considered urban.)
3. typology: Urban means places where housing units are older and located more densely than the average for the entire metropolitan area; suburban means where housing units are newer and more widely spread. (By this definition, over 80 percent of Americans live in suburbs.) (That's a lot.)

A more finely-grained approach is going to allow us to see with more clarity the strength and duration of the apparent move "back to the city" that began about the middle of the last decade. (William H. Frey at the Brookings Institution is one among many skeptics.) It also enables us to imagine what combination of features might be most attractive to new residents, while also being financially and environmentally sustainable. (More on this soon.)

Uptown Marion: not in the principal city, but arguably urban

For now, here is how the three approaches define city and suburbs in the Cedar Rapids context. The data come from the U.S. Census Bureau via this website, which is way easier to navigate than the Bureau's own site.

1. census-convenient: The City of Cedar Rapids is urban, with the rest of Linn County as suburban. The Census Bureau's Cedar Rapids Metropolitan Statistical Area (MSA) also includes Benton and Jones counties. That produces an urban population (2018) of 131,369, with 90,761 suburbanites in Linn County (136,955 suburbanites including the other two counties). The Cedar Rapids metropolitan statistical area is 49 percent urban.
2. suburbanisms: In the Cedar Rapids metro area, 83.2 percent of workers commute alone by private car; 69.8 percent of housing consists of a single unit ("single-family detached"), 74 percent of housing units are owner-occupied. By these definitions, the 18 urban census tracts are
1. below average on all three dimensions: 2.07, 7, 13, 18, 19, 27, 29
2. below average on all except car commuting: 3, 5, 6, 8, 10.03, 11.02, 15, 26, 30.02
3. below average on all except single-family housing: 22, 25. Collectively this includes the urban core of Cedar Rapids, excepting areas west of 10th St NW between E and O Avenues, as well as south of 1st Avenue, and Wellington Heights above 15th St SE. Outside of the core it reaches as far north as Blairs Ferry Road NE and includes the older sections of Marion; the areas around Kenwood Park on Cedar Rapids's northeast side and Cherokee Park on the northwest side; north of Washington High School on the southeast side; and from Jones Park on the southwest side as far south as 76th Avenue SW around Kirkwood Community College. That produces an urban population of 71,275; the metro is 26.6 percent urban.
3. typology: The median housing unit in the Cedar Rapids metro area was built in the 1970s. There are 57 housing units per acre, because much of the three counties is farmland. By these definitions, the 20 urban census tracts are 4, 7, 8, 11.01, 12, 13, 14, 15, 16, 17, 18, 19, 22, 23, 24, 25, 26, 27, 28, and 29. This includes the urban core of Cedar Rapids, excepting only the area west of 10th St NW between O Avenue and the Cedar River. Outside of the core it includes the Puffer District adjacent to downtown Marion; the areas around Kenwood Park on Cedar Rapids's northeast side and Bever Park on the southeast side; north of Washington High School and the Rompot neighborhood on the southeast side; and the southwest side stopping at Jones Park. This yields an urban population of 72,691; the metro is 27.1 percent urban.

Wellington Heights neighborhood's Stop the Violence Picnic, 2019 (swiped from redmondpark.org)
Census tract 17 qualifies as urban under definition three (housing density 2400, median age 1939)
but not definition two (76.1% owner-occupied, 81% single-family).

The first definition includes a significant chunk of Cedar Rapids that doesn't qualify under the other definitions, while excluding parts of Marion that do. The second and third definitions agree on 11 tracts, but the second includes seven while excluding nine that qualify under the third definition. Since all of the city qualifies under the housing density part of the typologies definition, the disqualifier always is median age of housing. For example, census tract 30.02 includes a degree of rental housing around Kirkwood Community College unusual for the metro, but the median housing unit was built in 2001 (and 87.1 percent of commuters drive alone). 

The disqualifiers for the second list ("suburbanisms") vary for those who qualify under the typologies categories. All nine tracts have higher rates of single-family housing than the metro average; eight have higher rates of car commuting, and five have higher rates of owner-occupied housing. Four tracts miss on all three! These include census tract 4, northeast of downtown Marion, which is bounded by 7th Avenue, 31st Street, 29th Avenue, and 10th Street/Indian Creek Road. The tract has 2089 housing units per square mile, and the median year of construction is 1961. Its car commuting rate is 85 percent, 88.5 percent of residents own their own homes, and 87.5 percent of residents live in single-family housing.

Why does it matter how we draw the urban-suburban line? Public perception of the line certainly affects how specific areas of the metro are perceived, though I'm not sure the analysis-by-census-tract featured above is going to have much impact on public perception.

The very fluidity (and arguability) of these lines teach us that the terms "urban" and "suburban" are not binary, that they each cover a range of neighborhood types and contain a range of people. One thing we can agree on: political boundaries rarely reflect social realities.

"Welcome to Suburbia!"