The 2010 Census Pt. I: The State of St. Louis

Tweet about this on TwitterShare on Facebook0Share on Reddit0Share on LinkedIn0Print this pageEmail this to someone

There is one typical question frequently asked of any urbanist. Its permutations may differ, but few cocktail parties, power lunches, or bleary-eyed salons are without someone speculating on a magic bullet to save the city. In the case of St. Louis the magic bullet is simple: more people.

This is the first of a multi-part series to examine the 2010 United States Census Redistricting Data in the context of the past ten years and seek to understand where we are today, what we can learn, and how we can position ourselves for the next decade.

With a background in urban design, it would be natural to critique the suburban-style low-density development that has mushroomed throughout the urban fabric in the past twenty years, but instead, I will focus on people, the residents.

The city inevitably fails to be a city when it tries to “do suburbia”, but that is a topic for another conversation. What is pertinent is that the demand and market conditions for density do not exist at this point in the majority of St. Louis’s 79 neighborhoods. Urbanists can bemoan strip mall infill projects, auto-centric development, and rampant demolition, but until we reshape the economic and political reality of the city such concerns will not lead to action.

Initial reaction to the population losses of resurgent urban neighborhoods of the near-southside was shock. This quickly formed into a dogmatic thesis that smaller, wealthier families in redeveloped neighborhoods account for the 10-20% population decrease across the near-Southside.

What is more concerning is the continuing loss of population density. By using the census population counts and the area of the neighborhood, the continuing loss of population density becomes clear.

While many debate whether population declines are attributable to smaller families replacing larger families in revitalizing neighborhoods,  whatever the cause, the resultant decrease population density remains undisputable. Although lower–income neighborhoods typically require more intensive social service provision, the provision of basic services such as roads, water, sewer, electricity, and gas do not change with socio-economic status. Whether the loss is due to abandonment or two people living in a building previously inhabited as a four flat apartment, the cost of maintaining basic services will escalate with fewer inhabitants.

This condition becomes even more stark when public transportation is taken into account.  Several established benchmarks determine minimum population densities for effective transit service. For minimum service of one bus per hour on a half-mile route spacing, four occupied dwelling units per acre are required (source). Using the average household size in St. Louis of 2.2 persons per dwelling, almost a third of St. Louis’s residential neighborhoods have lost the density required for minimal bus service.

This map is based solely on residential density criteria, so neighborhoods such as Downtown, Midtown and the Central West End possess a sufficient employment base to counteract lower residential densities and anchor efficient transit service. These neighborhoods have been demarcated with heavy borderlines on the map. In addition, this represents the density of the neighborhood population spread across the geographic boundaries of the neighborhood; places with small pockets of density may be viable transit nodes but will not be represented on this map with a neighborhood-level analysis.

Caveats aside, if access to efficient transit is essential for an urban setting, there are still more than twenty neighborhoods that can no longer be classified as urban neighborhoods due to prolonged and severe loss of population density.  Urbanism and urban potential are predicated on density, not history. We have to understand the present reality of post-urbanized areas clearly before we can hope to successfully address the future of urbanism in St. Louis.

If smaller families are indeed replacing larger families in revitalizing neighborhoods, it is possible that there will be reduced demand for consistent transit usage as new arrivals choose to work from home, or commute by bicycle or car. This subject needs further investigation through survey and fare box revenue before it becomes more than conjecture.

Part II will examine three differing stories present in the 2010 census data and examine closely the explanation of population loss as a result of smaller families replacing larger families and single-family rehabs replacing multi-family residential units.

Tweet about this on TwitterShare on Facebook0Share on Reddit0Share on LinkedIn0Print this pageEmail this to someone
  • marigolds6

    Block level data is available on MSDIS now too.
    If anyone wants this at block level, I can process it this morning and have it available someone in an hour or so, but I have nowhere to post it.

    • Alex Ihnen

      Sure! You can send it to [email protected] if you like.

      • marigolds6

        Still 30 minutes left on the rather massive download 🙂 Once I get that done, the rest should be pretty easy. I will do the population change and transit density analysis for St Louis City, St Louis County, and Jackson County.

  • Are we missing something simple here to explain population loss? Just look in local paper, in past two years you have had pages of foreclosure notices in Post Dispatch daily, with the majority of postings city properties. If just half of them resulted in people leaving their houses wouldn’t that account for a lot of population loss? It would be interesting to plot data for 2000-2008 verse 2009-2010.

    If the foreclosure crises really is a major source of population loss, you could possibly city a much quicker recover in the next five years.

    • Alex Ihnen

      This could have an effect. Unfortunately, we do not have apples-to-apples data for each year, so we can’t break down the numbers any more than 2000 v. 2010. My initial thought is that this may be a factor, but not a leading factor. To know more, we would need not just the number of foreclosures in the City of St. Louis, but information on where people moved after leaving their homes. The people don’t disappear. It’s conceivable that the City gained population due to foreclosures – there were likely more foreclosures in the County than in the City due to it’s size and it’s possible that some moved to the City to rent as there are a higher percentage of rental properties in the City. That said, I don’t think that we’ll ever really know.

  • marigolds6

    I was working through this for Jackson County and realized… shouldn’t percent change in population be exactly equal to percent change in density?
    ((Pop10/Area) – (Pop00/Area))/(Pop00/Area) = ((Pop10 – Pop00)/Area) / (Pop00/Area) = (Pop10 – Pop00) / Pop00

  • Great discussion, but as others have noted there are some flaws. Doing some quick calculations:

    square mileage of STL city (land area): 61.9
    acres per sq. mi.: 640
    total city acreage: (640 x 61.9): 39616 acres
    acres of parkland (according to STL parks service): 3200
    total city acreage for possible residential construction (39616-3200): 36416 acres

    Then using the number above of 2.2 person/dwelling unit in the city and the current census showing 319,294 residents, we can calculate:

    avg. dwelling units per acre (319,294 / 36416 / 2.2): 3.985

    So, in conclusion, I would disagree and say that most areas of the city can still support hourly bus service. Keep in mind that the above numbers don’t consider areas used primarily as industrial or commercial either nor does it consider that STL has a high poverty rate leading to even more use/need of bus service. So most residential areas could support greater than even hourly service IMO.

  • Another important factor to consider when predicting transit use is autos per household – a fairly low-density neighborhood will have high transit use if no one has a car, which is why your source data, based on research from 1977, probably needs to be updated a bit for today’s 4-car households.

    Great post though.

    • Douglas Duckworth

      “a fairly low-density neighborhood will have high transit use if no one has a car”

      The scenario of low density and high individual transit use (essentially transit dependency) exists in poor depopulated areas where people cannot afford a car and almost no where else. Others who can choose are not going to move to a low density neighborhood with the expressed preference of not using a car because the form predicates car usage. They will self-select to a density neighborhood which supports low automobile usage. Urban form (design) has a strong relationship with modal split. Higher density walkable areas fosters transit usage. Low density suburban form brings the car. The reason we don’t have the ideal now would be since we’re doing the opposite through policy: free or cheap parking, strip malls everywhere, overly wide city streets, along with a structure of low regional and local traffic congestion.

      Our loss of density has a lot to do with our practice of suburbanizing urban form which brings insufficient population concentrations for transit. Following the Tiebout argument we are doing nothing to differentiate ourselves from other suburban jurisdictions. We are giving people few reasons to vote with their feet back to urbanity because we are not creating it but in fact undermining our traditional form.

      Raising the cost or need to have a car through more compact and walkable urban forms will increase transit usage, especially if we are able to roll out transit systems at the same time which are seen as a replacement to driving. The problem being that this won’t happen absent government action on the scale of urban renewal and highway building. This does not preclude zoning reform to stop the horrendous practice of increasing the development cost of density or lower parking through the requirement of a zoning variance. We are sending the wrong signals or rather standards to the market.

      “The city inevitably fails to be a city when it tries to “do suburbia”, but that is a topic for another conversation.”

      “Urbanists can bemoan strip mall infill projects, auto-centric development, and rampant demolition, but until we reshape the economic and political reality of the city such concerns will not lead to action.”

      If we’re making a normative argument of how our city should progress (that we want transit, people, and density) then I don’t understand the purpose of this statement. Given academic evidence on this topic the only policy conclusion from this post is that the City — and Region — must entirely reform the planning regime. Perhaps that is the subject of the next post in the series?

      Regionally this could arise through merger and consolidation, State mandates, or inter-local agreements with teeth as in Denver. I think it needs to happen on the State and Regional level simultaneously for the biggest impact. The key being organized grassroots organizations statewide pushing for these reforms as what happen in Oregon. We need elected officials pushing for change in the legislatures with people on the street keeping that on their agenda. This is why Wisconsin Democrats felt emboldened by the grassroots to utilize the democratic process and relocate temporarily to Illinois.

      Ironically exurban development should foster smart growth constituencies outside the urban core. In Toronto the potential destruction of the Oak Ridges Moraine by sprawl actually crystallized exurban voters against further development which they saw as threatening their “rural” amenities. The Progressive Conservative (comparable to US Republican Party) government supported regulations to protect the ORM despite their traditionally strong ideological opposition to any interference which prevented outward sprawl. People can be pushed to compromise their positions if it’s necessary for them to win an upcoming election (Obama and triangulation).

      “We have to understand the present reality of post-urbanized areas clearly before we can hope to successfully address the future of urbanism in St. Louis.”

      Are you advocating planned shrinkage or reurbanization? I think it’s entirely fatalistic to support the former when the latter has had no policies — or even advocates of any political influence — promoting the proliferation of incremental increases in density which did occur in a few areas like downtown. If regionally we had a planning body with control over land use and transportation planning, while local jurisdictions were forced to conform to upper level plans, then perhaps the situation would be different. At least we need a City Planning Director!

      The problem with post-War American planning has always been localism, fragmentation, a willingness to talk about it yet no desire to abdicate some autonomy for the greater Regional public interest. Call it a problem with game theory or perhaps Ayn Rand gone amok. Perhaps we’re stuck in neoliberal capture or post-modern relativism where visions of the good no longer hold any validity. I reject the latter. Or am I spending too much time looking at governance models which have no potential in St. Louis? As customary now is the time to yet again dust off the Regional reform arguments. Agreeing with the Mayor for once I think that can be the only conclusion of the Census decline, while urban form remains imperative for the basis of speaking on reform. It is the key for our growth evident in our density increases downtown.

  • David

    Where did you get your data and graphics? I would love to look at similar trends in Kansas City.

    • Alex Ihnen

      Data is from and images were created by Andrew using GIS software, a lot of time and several pounds of coffee.

      • David

        They look very good, he did a great job. I’m having a hard time finding the actual redistricting data for those specific of locations. All I can find is county and larger. Any advice?

    • marigolds6

      I have similar maps put together for Jackson County, put nowhere to post them at the moment. Disqus image posting seems not to handle them. I will email to Alex and Andrew and maybe they can use them.

  • Malbrite10

    Since transit operates on roads, and roads serve all residential acreage in the city, a better measure would be density by specifically residential acreage.Take a look at two red, “hourly service infeasible”– Bevo, which has several large cemeteries and that huge shopping cluster along Kingshighway/Christy, and Tower Grove South, which is bisected by the Missouri Pacific Railroad and its associated industrial uses. Residential density in these neighborhoods is understated due to their high concentration of non-residential acreage and their large land area in comparison to other neighborhoods.

    It’s difficult to believe that Holly Hills is more dense than Bevo (sure it has probably more children/families – not exactly the heaviest transit users, by the way). Holly Hills–and Visitation Park, McKinley Heights, etc.–are mostly residential, small-ish neighborhoods though.

    The tale of Kings Oak is pretty cautionary to the above analysis. Its literal three blocks of housing would make hourly transit service feasible but the entire 1.5 sq. mile expanse of Bevo is not–even though a Kings Oak block is not dissimilar to a Bevo (residential) block in density.

    Given that, I would be careful in using this metric in declaring neighborhood “urban” or “not urban”. But all of your major points remain true. At a block group or census tract level analysis anyway.

    • Zun1026

      I agree that there needs to be some revision to this analysis. A look at the block group level would be more helpful. I did however like the overall direction of the article and the choice to examine STL from a transit supportive density perspective.

      • Andrew J Faulkner

        Unfortunately the 2010 redistricting data is at tract-level only and covers population, race, and units. Block group-level data will not be released for 6-9 more months. That being said, it is somewhat unclear how the city parsed the tract data to the neighborhood level and there are some anomalies such as a population gain in St. Louis Place that bear further investigation.

        • marigolds6

          The VTD data is census block level. It does not say so explicitly, but that is what it actually is.
          VTD data is posted to MSDIS now:

          That will give you block level information.

        • marigolds6

          Andrew and Alex. I just sent you both some reworked maps using the VTD data. In St Louis County, VTD=Census Block, but I don’t think the same is true for St Louis City.
          It is still a lower level of aggregation though.

          • marigolds6

            Second try at bus density map. Still failing 😛

  • Anonymous

    Can we re-map neighborhoods to break industrial areas out of residential areas? Does data like this exist, block by block. It would be interesting to see population and density maps like that.

    • Andrew J Faulkner

      There are datasets of landuse by parcel, but the most recent publically accessible set I am aware of is 5 years old. Additionally, landuses on paper rarely conform to actual use on the ground and fail to take into account illegal occupancy, covert loft conversions etc.