Understanding the carless
Most results reported in this paper come from analyzing data from the 2012 CHTS, which gathered geographically specific travel information from households in all of California’s 58 counties. The data were collected using diaries, computer-assisted telephone interviews, a website, and global positioning system devices. A total of 42,431 households recorded their travel for a pre-assigned 24-hour period, and provided detailed socioeconomic characteristics, locations, and household car ownership status.
To understand whether carless households chose to live without cars voluntarily, we analyzed the CHTS question that asks reasons for not owning a motor vehicle. Respondents who selected either “want to be without a car” or “concerned about [cars’] impact on [the] environment” (items 1 and 2 in Table 1) were assumed to have chosen voluntarily to forgo vehicles, provided they did not select any other answer suggesting that their choice was constrained (e.g., “for monetary, medical, or age reasons”). Conversely, households who could not afford vehicles, could not get insurance, or who had health- or age-related constraints, were deemed involuntarily carless, provided that they did not also give reasons that characterize voluntarily carless households. All other households were deemed “unclassifiable.”
Table 1: Classification of carless households
Following published car ownership studies, we investigated characteristic patterns of both households (income, education level, household composition, and dwelling type) and household heads (Hispanic or Latino status, other ethnicity indicators, age, gender, and immigration status). Since the built environment is an important determinant of car ownership, we also examined factors related to home area population density, land use diversity, and urban design. The nearby availability of public transit was estimated by measuring the share of the regional population that could be reached within 45 minutes via public transit and/or walking. Finally, residential self-selection — the likelihood that carless families choose to live in areas conducive to walking, bicycling, or riding transit — was statistically accounted for as well.
Who are California’s carless?
Households are more likely to be carless if they are African American, less educated, immigrated to California fewer than five years ago, or have many household members. Conversely, households are less likely to be carless if they have higher incomes, belong to the “Silent Generation” (born between 1920 and 1940), or live in a single-family house. Not surprisingly, compared to households with vehicles, carless households tend to live in denser, more land-use diverse, and more walkable areas with better transit service.
Similar factors are associated with households who voluntarily forgo their cars, with a few notable differences. Those with male heads of household are more likely to be voluntarily carless, but neither the age of householders nor the jobs-housing balance in their communities affects the likelihood of voluntary carlessness. Compared to involuntarily carless households, voluntarily carless households are also more affluent and live in more walkable, land-use diverse areas with better transit service. These differences between voluntary and involuntary carlessness suggest that the long-held practice of interpreting not having a car as an indicator of disadvantage without accounting for the fact that some people choose not to own cars may be distorting our understanding of household transportation decision-making.
Although its effects are relatively minor, our results also confirm the presence of residential self-selection — whereby some people choose to live in neighborhoods (such as central city areas) that make it easier to live without a car. Finally, while we find that higher population densities foster more voluntary carlessness, research has long shown that density alone has a relatively small influence on driving, and that other factors often associated with density — such as transit service coverage, neighborhood walkability, and the diversity of local land uses — play important roles in supporting carlessness as well.
Travel patterns of carless households
Figure 1. Linked trip frequency by distance
Figure 2. Linked trip frequency by trip duration
Compared to motorized households, carless households take fewer than half as many trips on average. Their trips are also consistently shorter; median trip distances are less than half as long, with a higher proportion of trips under a mile (Figure 1). Conversely, median travel times tend to be longer for involuntarily carless households than for either voluntary carless households or, especially, motorized households, as walking, getting rides from others, or taking public transit are typically slower than driving (Figure 2). Unsurprisingly, members of carless households ride transit, walk, and bike more than motorized households.
Compared to the voluntarily carless, involuntarily carless households travel farther, albeit less frequently, even though they are slightly less affluent. This is especially the case for trips involving personal business or work, civic, recreational, religious, or social activities. This may be because voluntarily carless households can satisfy more of their needs without traveling as far as they are more likely to live in neighborhoods with mixed land uses that make walking trips easy and convenient. Overall, voluntarily carless households walk and bike more, and depend on motor vehicles and transit less than involuntarily carless households.
The more frequent use of public transit by involuntarily carless households suggests that public transit still largely serves “captive riders” (i.e., riders without good alternatives to public transit), even though recent public transit investments have tended to focus on attracting so-called “choice riders.” Finally, when involuntarily carless household members do use motor vehicles, they tend to carpool more than when those in voluntarily carless households travel by car. Involuntarily carless households may have less flexibility when using motor vehicles because they are less affluent, on average, and thus have more incentive to share vehicles and rides.