Transit Tax Advocates Downplay Driverless Options
State Senator Scott Wiener, who recently authored a hefty transit sales tax bill for five Northern California counties, made his objections to robotaxis known to his followers on X.
— Senator Scott Wiener (@Scott_Wiener) September 24, 2025
But the tweet and the Bloomberg article it references do not fully come to grips with the impact of driverless technology and its potential to replace transit as we know it.
Transit advocates like Wiener often argue that because Autonomous Vehicles (AVs) typically carry a single passenger, they contribute to congestion. Buses and trains, by contrast, can accommodate dozens or even hundreds of passengers, taking many vehicles off the road.
But this is true only to the extent that most of the seats and standing positions offered by transit vehicles are occupied. And this is very often not the case, especially at off-peak travel times and in less dense areas. My analysis of June 2025 data from the National Transit Database shows that many bus systems across California (including those in Santa Clara County, San Mateo County, Central Contras Costa County, Riverside, and Sacramento) serve fewer than 20 passengers per vehicle revenue hour. These are systemwide monthly averages. And since the typical bus trip is normally much less than one hour, the number of passengers riding the bus at any given time is, on average, just a fraction of those <20 hourly riders.
Driverless vehicles can serve passenger loads approaching and ultimately surpassing the number of riders now using buses. Both Waymo and Zoox AVs can accommodate up to four passengers. Once these services implement shared rides, as Uber already does, they will be transporting multiple passengers per vehicle more frequently. Shared rides will also address another criticism of robotaxis: their high cost. Fares should also come down as the infant industry works through its teething issues, and as services begin competing with one another. Further, it would often be cheaper for transit agencies to subsidize robotaxi rides for low-income passengers than to operate mostly empty buses.
Also on the horizon is the prospect of driverless buses, which I discussed in my recent California Policy Center study of transit costs. Aside from the driverless bus pilots in Japan and Jacksonville which I covered in the study, we’ve now learned that Singapore is also adopting the technology. WeRide recently began operating robobuses without a safety driver around a 12-minute circuit in a tourist zone. Next year, the city-state will partner with Pony.AI to provide three longer driverless bus routes in a residential area.
As a center of innovation and a center of high labor costs, California should be on the cutting edge of robobuses, as it is with driverless cars. The fact that California is not keeping up with Japan, Singapore, and Florida on this front is likely the result of union control of the transit-industrial-complex. This is perhaps why transit advocates like Senator Wiener are always seeking new revenue for transit rather than searching for opportunities to make mobility more cost-effective by introducing labor-saving technologies.
Marc Joffe is a Visiting Fellow at California Policy Center.