Downtown Anytown, USA is a challenging research subject. Vague terminology, idiosyncratic boundaries, and limited data availability have contributed to a disconnected and incomplete body of research on the contemporary downtown. A downtown’s geography is a particularly vexing matter; there is no formal or even consensus definition for downtown; they are not recognized by the government, the Census Bureau, or the Postal Service, so many traditional sources for demographic and housing data are not readily applicable to a city’s urban core. Nearly all downtown research generalizes downtown districts to an aggregation of census tracts. The methods for identifying downtown tracts vary among researchers but notable methods include:
- Identifying tracts within a local downtown boundary
- Identifying tracts with centroids in an x-mile radius of city hall or of some antiquated Central Business District
- Identifying tracts within an area of high job densities
Each of these methods has strengths, but also major weaknesses:
- Local boundaries diminish the ability to compare across cities
- Uniform radius applied across cities ignores variations in size and topography
- Job density definitions assume that downtown must continue its historic role as a central business district (where it may in fact be primarily a cultural destination or residential district)
Arguably the greatest shortcoming of all of these approaches is their reliance on the census tract. No major studies have defined downtown without generalizing the boundaries using tracts. The reasons for this are straightforward: local population, housing, and workforce data is often derived from national sources like the Census and the American Community Survey. These sources only release limited data for geographies smaller than the tract, and for many downtowns tract definitions result in overgeneralization and would benefit from evaluation measures derived from smaller geographies.
Downtowns that do not have a major residential population contain fewer census tracts that cover larger geographic areas. Downtown boundaries built from these coarse geographic units may have little in common with local boundaries and can result in gross misrepresentations of downtown’s contents.
Tract-based downtowns are too generalized for the small, rapidly changing downtowns most in need of clear indicators of progress and precise measures of change. Perhaps a reason previous studies have not used this level of geographic precision is lack of awareness of sources for data at finer geographic scales. Some of the sources providing this level of detail are relatively new and trends in “open data” have made resources increasingly accessible.
The table below identifies some data sources that provide reliable and pertinent information at a geographic scale appropriate for most downtowns.
|Name||Smallest Geography Available||Data Provided|
(Longitudinal Employer-Household Dynamics Data)
|Custom areas (based on block units)||(For the employed population) Count, occupation, income brackets, ethnicity/race, distance and direction traveled from home to work|
|Esri Business Analyst Online||Custom areas (business data based on address, demographic data based on census geographies)||Businesses by industry, demographic and housing (derived from Census data)|
|Reference USA||Address||Businesses, including some historical|
|Local Tax Parcel data||Parcel (Address)||Property values, land use|
|CoStar||Address||Rent, vacancy, value, land-use, square footage, building history|
|Decennial Census||Block (much more data available for larger areas)||Population, ethnicity/race, household count, household type|
About the Author: Rachel Atkinson is a 2015 graduate of UNC Chapel Hill with a degree in her self-designed major: Urban Planning & Sustainable Development. Her thesis research dealt with the development of comprehensive downtown performance indicators. You can see more of Rachel’s work at www.rachel-atkinson.com. Currently, Rachel is working to launch Native South Creamery, making local, pecan-based milk alternatives.