How might travelers behave with privately-owned AVs?

By Michael Brenneis

In many ways, we can only speculate about a future with autonomous vehicles on the road. The effects on vehicle miles traveled are expected to be very different if AVs are privately owned versus shared. A recent post by Aaron Gordon on Jalopnick drew our attention to a still-relevant study, published in 2018, that focuses on the question of individual ownership. The study provided chauffeurs to 13 San Francisco Bay area participants for up to 60 hours over a 7-day period, and looked at the subjects’ behavior with, and without, the chauffeur-driven vehicle. The results point to a potential worst-case-scenario; a catastrophic increase in VMT that could occur with the introduction of privately-owned AVs.

The study followed the participants for three weeks: a chauffeur week sandwiched by normal weeks. The subjects included four millennials, four families (three with kids), and five retirees. The chauffeurs were professional drivers drawn from a service that drives clients’ own vehicles. While there was some variation in income level, age, gender, and household size, most of the primary participants had a college degree. The authors see the small sample size in this study as a beta test for a larger study.

In comparing the chauffeured week to the average of the pre- and post-chauffeured weeks, the researchers found that “VMT increased for 85 percent of the subjects (by amounts ranging from 4 percent to 341 percent), and the total VMT from the sample increased by 83 percent overall.” The lowered burden of driving, or possibly the novelty of having a chauffeur, led participants to make 58 percent more trips, take 91 percent additional longer trips (> 20 miles), and take 88 percent more trips in the evenings.

The subjects sent their chauffeured cars out on errands, to look for parking, or back home to wait. They allowed their friends to be chauffeured without them, and they sent their kids out without a parent. Driving without the primary participant accounted for 34 percent of the total increase in VMT—61 percent of that with only the chauffeur in the car. They took trips in the evenings when they might not normally, due to safety concerns. They took longer trips than they might have otherwise.

Interestingly, average walking miles went up for the retirees and families, but down for the millennials, although with some variation among individuals in the age cohorts. The authors speculate that walking may have decreased for some because some walking trips were replaced by driving trips, or because there was no need to walk to and from a parked car. When walking went up, it may have been because participants were taking more car trips (with some inherent walking) and were more active overall.

The authors point to the novelty factor, the lack of spontaneity given the 60-hour limit, and the presence of another human in the car during driving, as reasons that chauffeured cars may be an imperfect proxy for fully-autonomous, privately-owned vehicles. Given the last two, however, an even greater increase in VMT seems plausible. As Aaron Gordon points out, this study also doesn’t address the possibility that AVs could perpetuate sprawl and allow people to live even farther from the places they need to go.

Michael Brenneis is an Associate Researcher at SSTI.