Design Your Own Dream Subway Line – ConnectSF from City of San Francisco
connectsf.org
connectsf.org
By the time SF gets around to building any new subway line (10-15 years), private autonomous cars and shuttles will have siphoned away most of SF MTA's revenue. It'll be a lost cause. They need to build and _open_ a couple of new commuter subway lines, e.g. along the Geary corridor, before autonomous driving comes. Otherwise they'll struggle just to keep the existing bus and at-grade light rail networks working.
The irony is that roughly half or more of the budget for the Van Ness and Geary BRT lines is allocated for repaving the streets. They're effectively bankrolling their own death.
Madrid and other cities figured out how to build subways on the cheap. Too many U.S. cities see subways as prestige projects, and blow insane amounts of money on cavernous stations and other aspects of design and construction. If they built tunnels and stations like they often did 100 years ago--which is to say, as utilitarian public infrastructure projects--it'd be much cheaper.
How will a city that's already jam-packed with cars and losing fast thoroughfares to transit-only lanes be fixed with autonomous cars. You may not have to drive or park, but it will take 30-40 minutes to get around the city. A subway line would reduce this time in half which will always be appealing to riders.
I'm not disputing that subways would be preferable. I'm saying that from a fiscal and political standpoint SF MTA will never be able to justify and afford the capital investment unless they start building them RIGHT NOW.
The biggest cost in running a bus service is labor. Out-of-the-gate, autonomous cars and shuttles will have a substantially lower operating cost than the public bus service network, and will for some time. Middle-income commuters will flock to private bus services. Even if slightly more expensive, they'll be cleaner, more comfortable, and more timely, _especially_ for commuters. SF MTAs current fare is $2.25. A ride on Chariot (which I've used a couple of times) is less than $4. Imagine what Chariot could offer if they didn't incur labor costs for each shuttle?
The private shuttles don't obviate the need for a public bus service, but they'll absolutely take a huge chunk of the profits. Also, without middle-income commuters you're left with a higher proportion of people traveling on free or subsidized bus fares. Also also, average passenger count will decrease dramatically when you take away some of those commuters. An outsized proportion of revenue comes from packed buses, which if we're being honest is probably one reason why SF MTA has been slow with service upgrades--service upgrades not only cost money to implement, they reduce your margins.
San Francisco doesn't build subways because 1) it's too costly and 2) NIMBYs put up too many political roadblocks. Serious competition will mean even 1) less revenue for SF MTA, and 2) more excuses for the NIMBYs to block investment. Also, autonomous vehicles will only bolster lobbying for investment in the road networks at the expense of public transit.
The pain of slow transit _manifestly_ isn't enough to cause SF to build a new subway (the Central Subway notwithstanding for obvious reasons). Autonomous vehicles will only tip the balance in favor of surface transit generally, and away from public transit. It will make subways even more expensive in absolute and relative terms, which means they simply won't happen because (for good reason) SF has generally been a fiscally conservative government. The few recent big projects have come with substantial problems, despite being primarily funded privately or without state or federal money, and so will only make the city more risk-averse.
Long-term subways would still make sense. But from a long-term perspective subways have made sense in SF for quite some time. If investment can't be sustained from at least a medium-term perspective, it simply won't happen without a miracle. And autonomous vehicles will make the numbers look much worse than they ever have.