Will Autonomous Vehicles Mean the End of Public Transit?

By Leah Kaplan

Autonomous vehicles will mean the end of public transportation…or will they?

Autonomous vehicles (AVs), sometimes referred to as driverless or self-driving cars, have the potential to dramatically reshape our transportation systems. Some of these changes may be positive; AVs could improve road safety by reducing accidents caused by human errors and could open up new transportation options for individuals with disabilities. One significant area of concern, however, is how AVs might compete with other transportation modes, including public transit.

Transit systems have already had to reckon with competition from ride-hailing services (e.g., Uber, Lyft) that offer greater flexibility and convenience than many transit options. If autonomous ride-hailing services are significantly cheaper and more convenient than transit services, they could undercut public transit, which would have important implications for the environment and transportation equity. Public transit plays a critical role in reducing emissions from transportation, mitigating road congestion, and providing basic mobility for individuals with limited to no other transportation options.

The potential impacts of AVs, positive or negative, ultimately depends on one big question—do people even want to use them?

In our recently published research study, we conducted an online survey to better understand people’s preferences for different autonomous and non-autonomous transportation modes. In essence, we wanted to understand the conditions under which someone would get in an AV instead of taking public transit.

We used a specific type of survey called a choice-based conjoint survey that allowed us to explore a range of potential options that both currently exist and that could exist in the future. The image below shows one example of a choice question from our survey. Survey respondents were asked to imagine they were going out for an evening leisure activity and to choose their desired option, just as they might do if comparing options using a trip planning app. Each person answered 8 of these questions, but the features for each option (e.g., price, trip time) varied randomly for each of the questions so no person saw the same set of options twice.

One important feature that we included in our survey was having an onboard attendant. At the beginning of the survey, we explained that the onboard attendant would be a worker who isn’t operating the vehicle but who could provide in-person support. An onboard attendant could, for instance, help an elderly individual into or out of the vehicle.

Surprisingly, we found that people would pay extra to have a human attendant onboard their driverless ride-hailing vehicle. This may be because they don’t trust the AV technology or perhaps they really need that in-person support. Either way, the extra cost of paying for an attendant would mean that driverless ride-hailing vehicles could not out-compete public transportation. We also identified gender differences in preferences for AVs. Men were willing to pay more for automation and an attendant.

Using people’s responses to the choice questions, we were also able to estimate models for future marketplace competition between AVs and public transit modes. We found that for trips with good transit options, transit remains fairly competitive against AV modes, even with additional features like having an onboard attendant and a price discount. If ride-hailing was already more popular for a certain type of trip—perhaps because taking public transit would take significantly longer—it became even more popular with automation. This finding supports the idea that autonomous ride-hailing could help to fill existing transportation gaps.

With the continued development and gradual deployment of autonomous vehicles, AV companies and transportation planners will need to understand how quickly and willing the public is to emerging AV services. Gaining a greater understanding of public preferences for AVs enables transportation planners to design transportation systems that account for shifting preferences while still providing critical public transit services. AV firms can also use our study’s findings to help make important design and service decisions, such as whether to include an onboard attendant and at what price to set their service. At present, keeping human attendants onboard appears critical for both men and women, though these preferences could change as AV deployment expands and users gain more experience with these systems.