In Search of Common Definition of “Urban” and “Suburban” St. Louis

Tweet about this on TwitterShare on Facebook0Share on Reddit0Print this pageEmail this to someone

What is a city? What is urban? What part of the St. Louis area is urban? The United States Census Bureau defines “urban” and “rural”, but not “suburban”. And so the Census defines 81% of the U.S. population as living in urbanized areas, that is, everyone who doesn’t live in a rural setting.

In St. Louis, we sometimes bemoan a suburban development in the city, or an inner ring suburb. Sometimes “urban” equates to historic, or old: the City is urban, the County is suburban. Sometimes city political boundaries are used to approximate urban boundaries (everything outside St. Louis City is a suburb, or suburban).

As much as practicable, a standard definition of urban and suburban would be quite useful. The chief economist of the online real estate site Trulia recently took on the task (story at FiveThirtyEightEconomics). In short, they surveyed about 2,000 adults across the nation, asking them to describe their home as being located in an urban or suburban area, without prompting them with a definition of either.

The short answer on the result: 26% of respondents said “urban”, 53% replied “suburban”, and 21% “rural”. The 53% + 21% isn’t too different that the 81% “urbanized area” found by the Census. Interestingly, of respondents residing in cities of more than 100K residents, in metro areas of more than 500K residents, 56% responded “urban” and 42% suburban.

The short conclusion: a large number of Americans living in large cities describe their neighborhood as “suburban”. Researchers found that residential density of a respondent’s ZIP code was the best predictor of “urban” v. “suburban” response. Further, they found that a residential density of more than 2,213 households per square mile was considered “urban”, 102 to 2,213 “suburban”, and fewer than 102 households per square mile, “rural”.

Caveats: What’s considered “urban” in one city may be considered “suburban” in another. Definitions vary as context varies. Is an area relatively dense? It is relatively new? Residents of even dense small cities rarely described their neighborhood as “urban”. Conversely, residents of older, even lower density, areas often described the area as “urban”. Low residential density areas with a dense business district might be “urban”, as might less dense neighborhoods adjacent to higher density neighborhoods. That seems pretty straight forward.

Now, applying the found thresholds to some American cities shows, perhaps, just or urban or suburban our cities may be. New York and Chicago are 100% urban. Phoenix and San Antonio: 30 and 35%. We should know this. Sprawling cities, or those that have annexed large suburban areas, are, well, suburban. Large parts of most American cities are suburban.

This matters when considering economic development, crime, transportation, and more. It’s rather central to understanding, proposing policy, and developing a city. So what does St. Louis look like? What parts of the area are urban and suburban by the cutoffs discovered by the survey?

The ZIP codes below in blue are above the 2,213 households per square mile “urban” threshold. The areas in green are between 102 and 2,213 (“suburban”) and yellow is less than 102 (“rural”). To my eyes, this closely aligns with my notion of urban and suburban St. Louis. My only caveat being that I consider St. Louis City’s central corridor and north side, at least to I-70, as “urban” due to its historical dense development pattern. It’s true that larger portions today have suburban residential density, yet the street grid and remaining homes and businesses are clearly urban in nature. And that development pattern points to a possible urban future in a way that a suburban development pattern does not.

Beyond the city, we see the communities of University City, Clayton, and Maplewood as “urban”. Otherwise, the whole of St. Louis County is “suburban” other than two corners of rural density. In your mind, does the Trulia definition largely conform to your perception of St. Louis? What doesn’t make sense? Why does the “urban”/”suburban” divide matter in St. Louis?

urban v suburban STL

Tweet about this on TwitterShare on Facebook0Share on Reddit0Print this pageEmail this to someone
  • matimal

    How about defining “urban” through function rather than form. That is, an urban place is that with the greatest diversity of activities, interactions, and transactions.

    • Yeah, I think urban design (function and form) begets urban geography. Whether using some esoteric formula to determine an area as “urban” or something more intuitive/sensory, it all starts with a density of purpose.

      You can have (and do see) urban design in a suburb, but that doesn’t make the suburb urban. The reverse is true too — if some developer builds a vinyl-sided enclave on the City’s near south side, that doesn’t suddenly make the urban city suburban.

      The functional density works in tandem with density of form to define an area as urban. Same with the ‘burbs!

      • matimal

        So, if you build an urban space in a previously non-urban area, it will fostering urban function in and of itself? The Chinese ghost cities will eventually become urban places? We can will urbanity into existence?

  • Presbyterian

    This is really interesting. A challenge with using the number of residents per zip code is that it includes a bias against non-residential urban. The 63110 zip code (the green spot surrounded by blue) includes the Grove, Shaw, Dogtown and McRee, but it also includes nearly two square miles of Forest Park as well as BJC and CORTEX. That throws off the numbers.

    Still, I like that Clayton, University City and Maplewood read as urban. I tend to think these parts of the County have a lot more in common with the City than with the more suburban areas of the County.

    • Mary Ruoff

      I grew up in Richmond Heights, the narrow strip between Clayton and Maplewood. I would consider it urban as well, the bulk of the residential area anyway. Probably the Galleria , Target etc. throws off the demographics. Indeed, since I was a kid a number of residences have been removed for infill retail development. So, yes, you have to apply common sense, you just can’t go by zip code data alone. That’s what the kids from the county Catholic schools did back in the day when they came to the West End, U City, Richmond Heights, etc. for parties–that was all “city.” My neighborhood friend who went to high school out at Visitation (I was in the city at Rosati-Kain) used to tease about “hot child in the city, running wild and looking pretty” back in the day. Not kidding! LOL. She’s still my friend and still living in Richmond Heights and still a lot of fun!

  • HawkSTL

    Nice article — it shows that numbers roughly correlate to perception. I wonder if population density should be the sole data input. As an example, Detroit has an urban footprint (number of buildings per square mile, mix of multi-family and single family, and alleys). But, Detroit would now classify as suburban or even rural in some zip codes given the population decline. Point, being the physical layout also seems to have a great impact on the perception.

    • John R

      ^ I find it interesting though that Detroit has the same population density as we do; things are a bit different there as our depopulation has been concentrated in the North Side while they’ve had a comparatively more equal distribution of population loss throughout the city.

  • John R

    Thought I’d mention that Dempster Holland, long-time area urbanist, passed away. I believe he commented here on occasion.

    • Presbyterian

      Sad to hear we lost Demspter. He was a great friend of St. Louis.

  • Philamazoo

    Excellent post and great concept. I suggest you get down to a more granular level with census tracts or even census blocks (as census blocks are the compontent pieces of census-designated urban areas). Zip codes are too big to precisely define the boundaries of urban and suburban. I don’t live in St. Louis but know that in some parts of town near Wash U, “urban” and “suburban” neighborhoods seem to bump right into each other.

    • Alex Ihnen

      Couldn’t agree more – would be a great research project to survey maybe 10,000 people to provide enough info to create more accurate thresholds, perhaps segment by region of the country, and even city age, etc.

  • STLEnginerd

    2213 seems pretty arbitrary. Seems like someone decided to set the threshold so that a particular community they don’t think of as urban, would not make the cut, or so that a community they do consider suburban would make the cut. Either way there is no hard line definition, and we should just accept that. There are qualities the define urban/suburban areas which have been discussed ad nauseum.

    Also you are adopting political boundaries when in fact downtown St. Charles, and Newtown as we’ll as parts of Brentwood, Webster groves, Kirkwood are definitely urban. There is no clear single metric that demarcates the line, just accept that.

    • Alex Ihnen

      Sometimes I wonder if people read the article before commenting, or whether my writing is simply unclear and not very good.

      The number isn’t arbitrary, or the whim of the researcher. It’s based on the survey respondents – what they described as “urban” and “suburban”. There’s more on the methodology at the link provided. Perhaps my shorthand didn’t make this measurement clear. I think the whole point is that there isn’t and will never be a set, agreed upon definition, but that a better understanding of the definitions people use for their own neighborhoods can be useful.

      • STLEnginerd

        Truthfully it was early in the morning when i read it and responded so you have a point. I’ll try to remember to have some caffeine first next time.

        I see your larger point about self segregation into particular classifications and it is interesting. I guess my point beyond wild accusations of journalistic integrity is something you know and have acknowledge already. That there is no clear line and segregation doesn’t do us a lot of good in whatever form it takes IMHO.

  • JZ71

    A better indicator than “households” would be dwelling units. Yes, north city would be more “urban” if it weren’t, in many parts, an urban wasteland – it’s hard to have a defined household in a vacant, boarded-up structure.

    My definition of urban versus suburban would probably boil down to a combination of street frontage (small, for single family homes), front setback (small or none), alleys and, ideally, multi-story, multi-family, mixed-use structures along major thoroughfares. Another part of the equation is the time that it takes to get out of the “city” (any city) to a rural area. If you’re in Chicago, it can easily take hours to escape the seemingly never-ending suburban sprawl. A third component would be a viable public transit system.

    But, bottom line, does it really matter how you “define” things? Crime and poverty is not exclusively urban, suburban or rural, nor is propserity, good schools or truly “safe” neighborhoods. Much like trying to define “gentrification” or “blight”, there are many shades of gray, unless you focus exclusively on the extremes.

    • Alex Ihnen

      I think it can matter in the sense that “urban” and “suburban” are set ideas for many of us – attached to notions of what belongs there and what works there. But perhaps a greater awareness that Maplewood is “urban” could impact transportation planning. Instead of a BRT line to a “suburb”, it’s a line to an urban node not far from major employment centers. The point is to reach an agreed upon definition, but rather to continually re-examine, challenge, and add context to our understanding of development patterns and the current state of St. Louis.