ACS Data Users Group

Hello Everyone,

I am working on a project to estimate the number of people living in downtown. One of the first goals is for downtown organizations to better understand and serve its residents (e.g., a downtown residents council). We will certainly use Census 2010 count as a base line but given the new developments and undercounts of some vulnerable population, we will have to find other creative ways (e.g. utilities) that will get us closer to those numbers. Does anyone know of strategies, examples or resources of a similar project?

Best,
Gustavo.

As a city planner my approach would be to start with the 2010 Census as a base. I would adjust the figures based on newly units constructed units found within your geographic area. You presumably can obtain this info from your local planning or building department. (The various states and localities are organized differently in this respect.) Be sure to adjust population from new units based on the distribution of bedroom sizes and include a vacancy rate. I have found here in Cambridge, MA that new apartment community developments appear to have a higher vacancy rate than the overall housing stock. I would further adjust the population based on counts of vulnerable subgroups by contacting shelter and social service providers. It is also possible that your city conducts an annual homeless survey. If you track that down you can get more detailed information.

While I do not disagree with the methodology I was wondering how you defined your original Downtown/CBD. Is there an established boundary? Do other regions have an established downtown?

Good question. Since I was offering advice to another user I am not how he would go about answering this question. If I were the one faced with this task I think I would have to approach this question on a case by case basis. I am not aware of general standard, like that for MSAs. Most larger city planning or economic development departments probably have a working definition of downtown in their respective location. This could be based on historical understanding of the area, current zoning, or an economic development zone of some type. Another source could be a local business improvement district, which typically have well defined boundaries. In the case of Cambridge we do have defined commercial districts based on zoning and historical development patterns.

All good suggestions. One caveat to consider is that building permits only serve as a proxy for new construction as few municipalities and counties track new completions. While this can be a good indicator of building activity during a strong housing market, it can also be very misleading during a down market. In some areas, developments that were permitted never even got past groundbreaking. And the lots are still vacant, though the building permit data says otherwise. Also, make sure that the permit data you are examining only includes new construction and excludes renovations. Next you may want to stratify the estimate of housing units by owner-occupied and rental housing. The reason being that household size and occupancy (vacancy) rates can differ by type of ownership. Which brings up the important issue of occupancy. Unless you subscribe to the Field of Dreams theory (build it and they will come), the current occupancy will be somewhat lower than the estimate of housing within the defined market area. As dated as it is, you will probably need to rely on the Census for owner-occupied housing vacancy rates, but should consult a private source, such as REIS, for rental vacancies (they have some free stuff available online).

Good suggestions. Yes, the DDA (Downtown Development Authority) has established a boundary for what it considers to be downtown. However, part of the conversation will be to verify the current boundary and decide if further adjustments are needed based on current and future development patterns.

I should have been clearer about how to count new units. If you do not have detailed local knowledge about development since the 2010 Census, rather than building permits issued, I would use certificates of occupancy for completed structures as the source of the data.

[Updated on 3/25/2015 7:50 AM]

Is there any chance the downtown area can be approximated using census tracts or PUMAs? Local estimates may be more robust than ACS 5-year estimates, but you may be able to use the ACS results to fill in the blanks on vacancy and household size trends as you're crafting your own estimates.

I think this is a good approach depending on what you are trying to accomplish. For example, if you are seeking to compare your downtown to others elsewhere in the nation by all means use the 5 year ACS so you are comparing apples to apples.

Yes, I have discussed with the project team the geography question and aligning their boundaries to census geographies (census tracts). I still yet to receive the final boundary specs but they seemed open to the idea of adjusting boundaries. As soon as I get the boundary specs for the project, I intend to see how it aligns with census. PUMAs are quite large (4 for the county including 1 PUMA for the City of Grand Rapids. My assumption at this point is that they will be less inclined to use that large of a geography for the project. I am also looking into certificates of occupancy as one proxy for occupancy.