Transportation Lifestyles: Getting Started with Mapping
We should be getting more mileage out of vehicle availability data.
Welcome to my annual discussion of transportation lifestyles. This year, instead of exploring national and state trends in vehicle availability, I’m going to share some lessons learned about using the American Community Survey’s vehicle availability data to get a start at mapping transportation lifestyles.
(The U.S. Census Bureau has taken a cold, hard look at its pandemic data collection results and declared the ACS’s 1-year estimates for 2020 to be “experimental.” I’m going to follow their lead and hold off, for now, on analyzing vehicle availability trends.)
I didn’t start exploring the ACS’s vehicle availability data because I had a burning desire to know which state was more car-oriented: Washington or Texas. My initial interest was to see if the ACS data was useful for mapping transportation lifestyles at the census tract level. And the reason why I wanted to map them was because I had the hypothesis that transportation lifestyles could be a very powerful tool for coordinating transportation and land use.
Everyone has a transportation lifestyle: it is simply how a person connects to the world and gets things done. One of the useful traits of transportation lifestyles is that both transportation and land use are baked into the definition:
How connect → Transportation (and Telecommunication)
World → Land Use
Get Things Done → Well-being
Transportation lifestyles are applicable at the personal level, and also at the level of the household and the neighborhood. The concept is less useful for larger areas. For cities and states, it is better to think of them as collections of smaller districts or neighborhoods, each supporting its own types of transportation lifestyles.
Most people live multimodal transportation lifestyles, which ACS vehicle availability data alone obviously cannot describe. This is especially true if we define, as I do, “modes” to include the complete range of vehicle types (personal cars, taxis, mass transit, motorized small vehicles, bicycles, on foot, etc.) and also telecommunications, the home delivery of goods and services, and getting rides from family and friends. In addition, vehicle availability data doesn’t provide any information on how often the vehicle is driven and for which trip purposes.
ACS data is useful for two reasons: 1) the economics of personal vehicles provide a strong incentive to use a car if one is available; and 2) different types of household vehicle availability tend to cluster at the census tract level. This second reason is the subject of this essay.
Method for Mapping ACS Data
Conducting data analysis without first acknowledging the larger policy context or goal is never a good idea. Although the full planning paradigm for transportation lifestyles is a subject for a future essay, the conceptual framework in brief can be defined by two principles: 1) All transportation lifestyles are valid, unless political leaders direct otherwise. 2) The job of a transportation and land use planner is to improve people’s lives by making it easier to connect to the world to get things done.*
My annual trends essays have been built around four transportation lifestyles: car-free households, singles with only one vehicle, families with only one vehicle, and those living “car-two+,” defined as all households with two or more vehicles. The ACS defines vehicle availability as a car, van, or small truck available at home for personal use. Singles are one-person households. Families are households with two-or-more persons. Under this definition of family, household members do not need to be related.
Analyzing one-person households separately is paramount, as explained more here. One-person households are a significant and growing demographic segment of American society. In most states, people living alone ranges from 25–30 percent of all households. Singles with only one car do have vehicle resources similar to car-two+ households; however, in car-oriented places this transportation lifestyle is fragile in ways similar to other one-car households. In addition, people who live alone make up a majority of car-free households nationally.
In addition to the ACS’s household size by vehicles available data (ACS Table B08201), the ACS’s pre-packaged table on workers per household by vehicles available (ACS Table B08203) may also be useful. Car-free households with workers and car-two+ households with workers are subsets of car-free and car-two+ households respectively. One-worker/one-car households include both the single worker with only one car and families with one worker and only one car. Keeping with four household types, the final type is households with two-or-more workers and only one car.
The method seeks a meaningful, but simple way to map vehicle availability. The maps shown here use only three categories, roughly translating into low, medium, or high concentrations of a lifestyle type in a census tract. Another way to interpret these categories are as answers to the question “Does this census tract support a critical mass of a specific lifestyle?” The three categories — low, medium, and high — would translate to “probably not,” “maybe,” and “yes.”
Maps of Phoenix Transportation Lifestyles
Although the fifth largest American city by population, Phoenix’s split of vehicle availability types is similar to the country as a whole. In both, a majority of households live car-two+. The most significant differences are that Phoenix households are more likely to be families with one car or households with one worker and one car (see Table 1).
Since mapping vehicle availability data is just intended to get a start on sketching transportation lifestyles at the neighborhood level, it usually makes sense to start with the lifestyles lived by a minority of households. Because those living the majority lifestyle are better positioned to shape society and infrastructure around their needs, those living minority lifestyles are also more likely to be on the losing end of any disparities in how easy it is to connect to the world and get things done.
Phoenix’s car-free households cluster in central Phoenix, although only nine census tracts have “high” critical mass (see Map 1).
The 33 census tracts with “high” concentrations of families living car-lite with only one car are a little more spread out (see Map 2). Census tracts with “medium” concentrations of families living car-lite actually outnumber census tracts with “low” concentrations (197 to 151).
One-person/one-car households (Map 3) and one-worker/one-car households (Map 4) have the largest numbers of high-concentration census tracts at 71 and 92 respectively. Because Phoenix has an unusually low percentage of households with no workers, one-worker/one-car households actually outnumber one-person/one-car households.
For the two minority lifestyle types not represented here by maps: Phoenix has no census tracts with “high” critical masses of car-free workers or two-or-more workers and only one car. “Medium” concentrations are also few in number, numbering 20 census tracts for car-free workers and 16 census tracts for workers with only one car.
A Tale of Two Transit Centers
The Census Transportation Planning Products can also be used to analyze and display the ACS’s vehicle availability data. Despite its somewhat unwieldy user interface, the way it presents data can be very useful. It is also an easy way to examine the vehicle availability of households with children.
To illustrate, I used the Census Transportation Planning Products to compare two transit centers, Downtown Silver Spring and Langley Park, located in Maryland near the Washington, DC border. Downtown Silver Spring is served by a metro station, commuter rail station, and a bus transfer center. The street network still benefits from the connectivity of a grid, and its mix of housing, services, and commercial development reflect decades of transit-oriented development policies. Langley Park, which has only a bus transfer center, is auto-oriented commercial and residential development dominated by parking lots and wide commercial arterials.
Downtown Silver Spring supports a critical mass of car-free living; it’s two central census tracts — at 855 and 1,255 households — easily exceed the “high” concentration threshold. At first, Langley Park seems undifferentiated from the dominant “medium” concentration of car-free lifestyles of the larger region. Looking at the actual numbers for its two central census tracts reveal that, at 495 and 410 households, they are approaching the “high” category. (A helpful reminder that analysis boundaries are always a little arbitrary.)
However, examining car-free households with workers begins to tell a different story. Downtown Silver Spring and Langley Park are likely to be roughly equivalent in their support of car-free workers, because Langley Park has more workers per car-free household. (See Map 5.)
For children living car-free, Silver Spring has almost none, while Langley Park hosts a “medium” concentration. (See Map 6.)
Examining one-car households again highlights the differences in lifestyles between Silver Spring and Langley Park. Because of the larger number of workers per household, Langley Park’s one-car households have more workers served by the household’s car. (See Map 7.)
Langley Park also has more children living in one-car households. (See Map 8.)
Already at this preliminary level of analysis, when examining how people connect to the world and gets things done, what is going on in Downtown Silver Spring appears very different from what is happening in Langley Park. An analyst would be on pretty solid ground hypothesizing that the multi-modal lifestyles in Langley Park are likely to be lot more complex and complicated than the multimodal lifestyles in Downtown Silver Spring.
A Note on Income
I know from experience that when I present these maps to live audiences someone either directly confronts me or starts mumbling with the observation that these maps are just maps of income. Places that “support” car-free and car-lite lifestyles are really just places inhabited by people with lower incomes, they argue. My three-part response is:
A) Yes, this is often true;
B) But this is not always true: moreover, the places that are exceptions are just as important as places that follow the general rule; and
C) Income is not really relevant, at least not at this point in the analysis. See Principles 1 and 2.
If all transportation lifestyles are presumed valid, my priority as a transportation and land use planner shouldn’t be divining why you are living a perfectly valid lifestyle. My priority should be building an understanding of how well your lifestyle is meeting your needs. And, ultimately, only you are an expert in how well your needs are being met.
Factors such as income and other sources of disparities in ease of connecting to the world such as race, immigration status, gender, age, and disability do become important — often very important — at the stage of developing strategies and tactics for interventions. These factors also may be relevant in developing strategies and tactics when faced with lifestyles that the powers-that-be have deemed — de facto or de jure — invalid.
At this stage in the analysis process, however, we’re not yet developing strategies and tactics. Indeed, developing strategies and tactics designed to make it easier for households and neighborhoods to connect to the world and get things done is for planning processes conducted under the guidance of political leaders and in partnership with community members.
Mapping lifestyles according to ACS availability data, on its own, is designed to raise questions, not provide final answers. This section discusses three potential uses for vehicle availability data mapped at the census tract level.
A note on method: I don’t foresee a future when it will be all that useful to outsource transportation lifestyle analysis to some third-party, data crunching shop. I doubt that there will ever be a magic equation or formula for transportation lifestyles. A complete analysis of transportation lifestyles would include an expert panel or other qualitative analysis processes that draw on those with a diverse range of local knowledge. The analysis may also lend itself well to a community science or citizen science project.
1. Analyze the patterns of lifestyle concentration and dispersal and begin to think through their implications. You may want to go beyond the four (or eight) starter lifestyles and see if additional permutations produce anything of value. If you are analyzing a city with a large number of car-free households, mapping singles, families, and families raising children separately may be illuminating. For a community with a large retired population, focusing on households with no workers may be of value. If there are neighborhoods with a large number of singles living car-two+, this is also an indicator of a very distinct lifestyle.
Dispersal may be just as important as concentration. Take for example a county where car-free households are roughly evenly split between the east and west halves. In the east, they are concentrated in specific neighborhoods, but in the west they are widely dispersed among all the neighborhoods. In the east, land use strategies would definitely be on the list of strategies worth considering when planning how to make it easier for car-free households to connect to the world. In the west, land use strategies would likely be a lower priority.
Finally, this may be an appropriate time to start looking for census tracts that are exceptions to the “income rule” that more income means more car-oriented lifestyles and for other unique or rare transportation lifestyles. It is important to emphasize here, however, that the analysis should be looking for transportation lifestyles that people are actually living, not a planner’s fantasy about the lifestyles that people could or should be living.
2. Use as an input when defining and designating neighborhoods for further analysis and for prioritizing planning and investment efforts. Data on car-free households is often used in the identification of disadvantaged neighborhoods, but we could be doing so much more.
3. Use as a starting point for what I’ve been calling (for lack of a better term) a “neighborhood transportation/land use portfolio.” In brief, this portfolio is a report that documents how — and how well — people in a neighborhood connect to the world to get things done. Although the final version of such a report should be driven by community input, an analyst or staff planner can get started on a preliminary draft by using vehicle availability data to define a set of lifestyle types for a neighborhood. By analyzing these lifestyle types with what amounts to empathy, common sense, basic transport and land use data, and shoe leather, a preliminary draft of what it would be like to try to live different transportation lifestyles in a specific neighborhood can be constructed.
Using ACS data or other data sources on car-free households is standard practice in some types of transportation planning and even some land use planning. I believe, however, as planners we should be getting more mileage out of the ACS vehicle availability data. My hope with this essay is to prompt others to think creatively about how to use this data, in pursuit of the larger goal of making it easier for people to connect to the world and get things done.**
Sarah Jo Peterson is the founding principal of 23 Urban Strategies, LLC, which works at the intersection of transportation, land use, and sustainability.
* Yes, this could be seen as radical redefinition of the job of the transportation and land use planner. I’m arguing that the job of helping people connect to the world and get things done is well served by expertise in both transportation and land use, as well as other related subjects.
** A note of appreciation to Takiya Louers for the maps of Phoenix and to Peter Volosin and David King for their interest, over the years, in engaging with me on the uses of vehicle availability data.