There’s No Such Thing As the Suburbs

For a recent report, TTI was faced with the challenge of classifying 1,300 survey response zip codes as urban, suburban, or rural. Researchers wanted to know where respondents lived to look at how shopping trends might affect freight and deliveries in different areas. We asked respondents things like “are you an Amazon prime member?” and “how many package deliveries did you receive in the last week?” We wanted to know, in part, where most deliveries were going. For example, should we expect tons of Amazon trucks making their way to suburban Texas? Or are they more likely to be double-parked in downtown Austin? [1]

The problem with trying to classify areas as urban, suburban, or rural is that, according to the U.S. Census, the suburbs don’t exist. Don’t get me wrong, places like Sugar Land, Texas (suburb of Houston), and Pasadena, California (suburb of Los Angeles), aren’t going to disappear off the map. It’s just that when the Census classifies geographic areas, there are only two categories: rural or urban. Aside from the Census, there are a number of ways to classify areas on an urban-rural scale. However, there is no commonly accepted method. Understanding where people live, and where they receive deliveries, is essential to invest in infrastructure where it is needed most. For those of us working with population and geography data, this lack of common definitions creates a challenge when trying to define an area as urban, suburban, or rural.

Let’s look at some of what’s out there.

The U.S. Census

According to the census, an area is “urban” if it has 2,500 or more inhabitants.[2] The census distinguishes between urbanized areas[3] (UAs) and urban places outside of UAs, but both are included in the definition of urban. The census has no definition of suburban areas.

Under this classification, cities as small as Hebbronville, Texas, are considered urban. Hebbronville (population 5,350) is located in rural South Texas, an hour west of Laredo. Hebbronville has a population density of 4.6 people per square mile, and a housing density of 2.2 households per square mile. For contrast, one of the central neighborhoods in Austin has a population density of 4,838 people per square mile and housing density of 2,708 households per square mile.[4]

United States Department of Agriculture (USDA)

The USDA uses a rural-urban continuum code (RUCC) by county.[5] Counties receive a score based on population size of metro areas within the county, or degree of urbanization and proximity to a metro area. There are three metro categories and six non-metro categories. Information on every county in the U.S., including population, rural-urban continuum code, and description can be found here.

Under this classification, Jim Hogg County (home of Hebbronville, Texas) receives a score of 7 — it is non-metro, has an urban population between 2,500 and 19,999 and is not adjacent to a metro area.

Fort Bend County (home of Sugar Land, Texas) and Harris County (home of Houston, Texas) both receive scores of 1 — they are counties in metro areas that have a county population of 1 million or more.

The Department of Defense

The Department of Defense created definitions for rural, urban, and suburban zip codes for the Medicare Modernization Act of 2003.[6] Information on population density by ZCTA code[7] can be found here.

Using this classification, Hebbronville is considered rural because it has fewer than 1,000 people per square mile. Sugar Land, Texas, a Houston suburb, is considered suburban with a population density of 2,543 people per square mile. Certain zip codes within Houston city limits would also be considered suburban. For example, the Houston Gardens neighborhood in Northeast Houston, with a population density of 1,832 people per square mile, is considered suburban, even though it is located only 10 miles from downtown Houston.

Trulia Survey

In a survey done by Jed Kolko, the former chief economist at Trulia who now holds the same position at Indeed, researchers asked participants for their zip code and whether they thought they lived in an urban, suburban, or rural area. Their reason for doing so was to capture residents’ own experiences of “urbanness,” as opposed to imposing outside definitions. Survey results showed that the best predictor for a respondent calling their area urban, suburban, or rural was housing density.[8] Therefore the study team created a model to code each ZCTA as urban, suburban or rural, with housing density as the main factor. A full explanation, as well as access to the final code book is located here.

Similar to the Department of Defense descriptions using population density, the Trulia survey classification considers Hebbronville rural and Sugar Land suburban. However, under this classification the Houston Gardens neighborhood (704 households per square mile) in Houston is considered urban even though it falls within the suburban housing density range listed above. This is because survey results showed that residents of low-income neighborhoods, such as Houston Gardens, often classified their area as urban, even if housing density was lower than 2,213.

Why Does This Matter?

Each of these methods for classifying areas as urban, suburban, or rural incorporate different information and have their own pros and cons. However, for TTI’s purposes in this study, we chose the method developed in the Trulia survey. We did not feel comfortable with the USDA rural-urban continuum code because it defined area at the county level — an area we felt was too broad to classify as one thing. And the Department of Defense definition uses population density to classify zip codes as opposed to housing density. We felt that housing density was a better indicator of the built environment. For example, if a neighborhood lost population density because single residents or couples moved into homes that were previously occupied by larger families, it wouldn’t necessarily be less urban.

Using Mr. Kolko’s classification system, we found that over 60 percent of our respondents live in the suburbs — a geographic definition not recognized by the U.S. Census. But why does this matter? Understanding where people live is crucial to investing efficiently and planning for the future. In the case of freight, knowing where deliveries are going could help both the public and private sector plan for the most efficient manner to accommodate trucks.

There is no perfect definition for urban, suburban, or rural areas. But having common definitions released by the Census would allow researchers across different institutions and fields to compare and debate their work more easily, both of which lead to idea sharing and richer understanding. A common definition also allows for tracking changes over time, without having to reassess or explain the areas each time. In the meantime, researchers working with geographic areas should be careful to disclose how they classified each area.

Without a consistent definition for some of the broadest and most populated spaces in the U.S. — the suburbs — we may lack understanding of where investment and services are needed most.

[1] Urban Area: a continuously built-up area with a population of 50,000 or more

[2] ZCTA is similar to zip code. To convert zip code to ZCTA use this.

[3] Image source: James Fenton, The Noun Project. Available at: https://thenounproject.com/.

[4] U.S. Department of Commerce, “Geographic Areas Reference Manual.” November 1994. Available at: https://www2.census.gov/geo/pdfs/reference/GARM/Ch12GARM.pdf, accessed February 2017.

[5] U.S. Census Bureau, American Fact Finder. Based on 2010 census data.

[6] United States Department of Agriculture, “Rural-Urban Continuum Codes.” Available at: https://www.ers.usda.gov/data-products/rural-urban-continuum-codes/, accessed February 2017.

[7] Federal Register “Part III, Department of Health and Human Services, Centers for Medicare & Medicaid Services, 42 CFR Parts 403 and 408 Medicare Program; Medicare Prescription Drug Discount Card; Interim Rule and Notice,” December 15, 2003. Available at: https://www.cms.gov/Regulations-and-Guidance/Regulations-and-Policies/QuarterlyProviderUpdates/downloads/cms4063ifc.pdf, accessed February 2017.

[8] Jed Kolko, “How Suburban Are Big American Cities?” FiveThirtyEight.com, May 21, 2015. Available at: https://fivethirtyeight.com/features/how-suburban-are-big-american-cities/, accessed February 2017.

Sarah Overmyer is an assistant transportation researcher at the Texas A&M Transportation Institute.

• Prior blog entries
• 85^th Regular Session Transportation Legislation
• TTI-PRC coverage of 84^th Interim
• Subscribe to Transport Trends newsletter