Glenn, thanks for the thoughtful response (I'm a fan of Car Charts by the way!)
I think the Silicon Valley myopia on efficiency is a fair critique. And I agree that personally owned AVs will be part of the picture.
Where I'd push back a little is on how static the framing is. We spent 3-4 generations making driving and storing your car hyper-convenient, with free parking everywhere and cities built around the assumption every adult drives. I'm not sure we will keep making those decisions once AV rides cost less than driving yourself. And generationally, the attachment to cars as identity is already fading. Millennials still feel it and Gen Z less so. By the time the cost lines on that Goldman Sachs chart cross (and that will surely happen at different times in different places), the generation coming of age might not care about car ownership the way we do.
I think the honest answer is it'll be really interesting to see how it plays out. The one thing I don't have a great answer for is induced demand. If rides get cheap enough, people take a lot more of them, and we're probably not thinking about that enough.
Your reasoned reply to my slightly-unhinged rant all makes sense. I will push back though on "attachment" to cars. (Clears throat, begins lecture along lines of "I had to walk to school every day in the snow when I was young, you kids have no idea.....") I am 70. In 1986 in one of my first McKinsey assignments we produced research showing Young People Are No Longer Attached To Cars As They Once Were. And I have seen reports repeating this every few years since. It's a little bit like classical orchestras bemoaning how elderly their audience is, such that in X years they'll have to fold up shop.... and every X years that elderly audience renews itself. Driver's license data shows young people DEFERRING getting their licenses more than they used to... but then again more are going to college or grad school than before, deferring getting married, deferring getting full-time work, etc.... but they eventually DO GET THEM. Thus the total count of driver's licenses in the USA continues to increase. Nothing like having a baby to get you out of the Uber and into your own car. But I digress (my middle name....). I don't think the "attachment" matters much. I agree the emotional connection is weaker - heck in 1970 if you wanted to have illicit High School sex the only safe haven was the back seat! This Has Changed. But I don't know many people who are "attached" to their dishwashing machines either, and they still keep buying them. My points are: adoption will be slower than we generally think, and a new equilibrium will be found. See my Car Charts post showing cars/household GROWING in San Francisco SINCE the launch of ridehail. Hmmm. Okay, time for me to go yell at the kids in the front yard and then take a nap.
Fair point, though every time someone said "young people don't care about cars" over the past 40 years, there was no viable alternative to owning one. Now there might be.
I do think the timeline is long. More of a 30 year thing than a 10 year thing. And maybe I just really want this to be true: I grew up in Kentucky where a car is both an identity statement and a precondition to participating in society, and addressing that is what I've dedicated my career to, first in public transit and now in AVs. I think AVs are the first technology that could actually change that equation, and I think it's really starting now.
Well, the same thing about a viable alternative to ownership was said about ridehail when it launched: "Uber CEO Kalanick says car ownership is becoming passe: 'Millennials aren't buying cars anymore. They don't want to drive. They don't want to own cars.'" (Newsweek, 2017) And since Uber launched I can't find a single metro area in the USA where cars/household has dropped in any meaningful way, and in many it has risen. But don't get me wrong: I think AVs will take off for sure, but I think in the long run they will be as much privately owned (Tensor) rather than rented (Waymo), or more likely, a mix of both. And Americans do have too many cars! My best GUESS is by 203X if we were headed for 2.5 cars per household, the rise of AVs (rented OR owned), combined with dwindling family size (nothing puts you in a hulking SUV faster than three kids with hockey equipment) will knock that down to 2.0. Maybe less. But all the feverish "My daughter will not need a driver's license" talk (remember who said that?) I think is overdone. I will now shut up!
Right, but Uber was never a viable alternative to ownership for most people. It was a convenience product for people who already had the option of driving. If AVs get to $1/mile or below, that's a different proposition. We'll see!
And I don't know who said "My daughter will not need a driver's license" but I do remember that sentiment in the first AV hype cycle ~7-8 years ago.
I agree with your take on Uber now, but remember at the time people like Kalanick (and ReThinkX, remember them?) asserted it WOULD be a viable replacement. And the quote was from Chris Urmson. (And it was about his son, I got that wrong.)
Thanks for the feedbakc. We moved the footnote to right below the chart so it's clear that Tesla is operating with safety drivers (in CA) / operators (in TX). The goal of this chart was to highlight the different services and where they're operating, and also what stage they're at without cluttering the actual chart too much. So hopefully this achieves that now.
Fair point. Tesla's operating with safety drivers and a waitlist, not driverless. Footnote 1 calls this out, but I should make it more visible. Thanks for the feedback.
This is a superb post with lots of info and analysis: bravo! And thanks. I will quibble a bit with the industry's assumption that once the cost/mile lines cross massive adoption will take off. (We said the same thing about TCO for EV versus ICE, remember?) I like robotaxis, I use them whenever I can, and I think their usage will soar. But I am not sure that the rosiest adoption scenarios will come to pass, because POINT TO POINT COST IS NOT THE ONLY THING THAT MATTERS TO CUSTOMERS. (If it were, no one would ever buy a new car in fact: a clean used Accord would beat any new car in cost/mile!) There are so many other sources of value in using your own car versus a rental/taxi/human-Uber/robotaxi: More CONVENIENT to jump in your own conveyance and go where you want when you want. We value our time and our capriciousness. Who wants to always be poking a phone to see where the darn cab is? COMFORT: You can set up your car to cater to your tastes, interests, needs and comfort. STUFF. Many of us treat our car as a backpack, full of stuff available to meet occasional needs: golf clubs, change of clothes for the gym, car seat for the kid. INDIVIDUALIZATION: Having your car as a statement of what kind of person you are can be a significant share of its value. Why does anyone ever buy a sports car? Arriving in an Uber says nothing. CONTROL: You can fix your costs of transportation with your own vehicle. Once I give up my own car, I am at the mercy of whatever prices the ridehail or robotaxi company sets. I can defer maintenance if I want, take the car to Cousin Fred for cheap work the OEM would never approve of. HEALTH: I can keep my car clean – or not, as I want, but I know who’s been in it: me. Do I want every trip I take to be with germs other riders have left behind? Remember Covid? RANGE: Ridehail and robotaxi economics depend on population density, reducing service and increasing price for rural trips. I want to be able to go anywhere at any time, with low marginal cost. ..... To me what this all says (plus falling robotaxi costs) is: PERSONALLY-OWNED AV. The AV tech is so great, and is getting cheaper, why not own the darn thing, versus have to summon-and-pay over and over? Put the kids' car seats in the back and leave 'em there. Leave the groceries in the trunk while you grab a ramen bowl: no need to summon a robotaxi twice. Etc. The future I see is: "Cadillac... plus Waymo Inside©" (grin) If cost/mile rules, why does ANYONE in Manhattan own a car? Because 25% of households there DO. My view: we find a new equilibrium: long trips (airplane), micro-trips (bike and feet (Shoeber)), medium unstructured trips (rental car), simple point-to-point trips (human or robo-ridehail), complex errand trips (owned car, AV or not AV). Etc. Silicon Valley always looks myopically at EFFICIENCY (doing things right) and not at EFFECTIVENESS (doing the right things). End of Rant!
Glenn, thanks for the thoughtful response (I'm a fan of Car Charts by the way!)
I think the Silicon Valley myopia on efficiency is a fair critique. And I agree that personally owned AVs will be part of the picture.
Where I'd push back a little is on how static the framing is. We spent 3-4 generations making driving and storing your car hyper-convenient, with free parking everywhere and cities built around the assumption every adult drives. I'm not sure we will keep making those decisions once AV rides cost less than driving yourself. And generationally, the attachment to cars as identity is already fading. Millennials still feel it and Gen Z less so. By the time the cost lines on that Goldman Sachs chart cross (and that will surely happen at different times in different places), the generation coming of age might not care about car ownership the way we do.
I think the honest answer is it'll be really interesting to see how it plays out. The one thing I don't have a great answer for is induced demand. If rides get cheap enough, people take a lot more of them, and we're probably not thinking about that enough.
Your reasoned reply to my slightly-unhinged rant all makes sense. I will push back though on "attachment" to cars. (Clears throat, begins lecture along lines of "I had to walk to school every day in the snow when I was young, you kids have no idea.....") I am 70. In 1986 in one of my first McKinsey assignments we produced research showing Young People Are No Longer Attached To Cars As They Once Were. And I have seen reports repeating this every few years since. It's a little bit like classical orchestras bemoaning how elderly their audience is, such that in X years they'll have to fold up shop.... and every X years that elderly audience renews itself. Driver's license data shows young people DEFERRING getting their licenses more than they used to... but then again more are going to college or grad school than before, deferring getting married, deferring getting full-time work, etc.... but they eventually DO GET THEM. Thus the total count of driver's licenses in the USA continues to increase. Nothing like having a baby to get you out of the Uber and into your own car. But I digress (my middle name....). I don't think the "attachment" matters much. I agree the emotional connection is weaker - heck in 1970 if you wanted to have illicit High School sex the only safe haven was the back seat! This Has Changed. But I don't know many people who are "attached" to their dishwashing machines either, and they still keep buying them. My points are: adoption will be slower than we generally think, and a new equilibrium will be found. See my Car Charts post showing cars/household GROWING in San Francisco SINCE the launch of ridehail. Hmmm. Okay, time for me to go yell at the kids in the front yard and then take a nap.
Fair point, though every time someone said "young people don't care about cars" over the past 40 years, there was no viable alternative to owning one. Now there might be.
I do think the timeline is long. More of a 30 year thing than a 10 year thing. And maybe I just really want this to be true: I grew up in Kentucky where a car is both an identity statement and a precondition to participating in society, and addressing that is what I've dedicated my career to, first in public transit and now in AVs. I think AVs are the first technology that could actually change that equation, and I think it's really starting now.
Well, the same thing about a viable alternative to ownership was said about ridehail when it launched: "Uber CEO Kalanick says car ownership is becoming passe: 'Millennials aren't buying cars anymore. They don't want to drive. They don't want to own cars.'" (Newsweek, 2017) And since Uber launched I can't find a single metro area in the USA where cars/household has dropped in any meaningful way, and in many it has risen. But don't get me wrong: I think AVs will take off for sure, but I think in the long run they will be as much privately owned (Tensor) rather than rented (Waymo), or more likely, a mix of both. And Americans do have too many cars! My best GUESS is by 203X if we were headed for 2.5 cars per household, the rise of AVs (rented OR owned), combined with dwindling family size (nothing puts you in a hulking SUV faster than three kids with hockey equipment) will knock that down to 2.0. Maybe less. But all the feverish "My daughter will not need a driver's license" talk (remember who said that?) I think is overdone. I will now shut up!
Right, but Uber was never a viable alternative to ownership for most people. It was a convenience product for people who already had the option of driving. If AVs get to $1/mile or below, that's a different proposition. We'll see!
And I don't know who said "My daughter will not need a driver's license" but I do remember that sentiment in the first AV hype cycle ~7-8 years ago.
I agree with your take on Uber now, but remember at the time people like Kalanick (and ReThinkX, remember them?) asserted it WOULD be a viable replacement. And the quote was from Chris Urmson. (And it was about his son, I got that wrong.)
Tesla is NOT doing AV in the SF Bay Area yet, nor even in Austin. Makes me seriously doubt any of the data in this post.
Thanks for the feedbakc. We moved the footnote to right below the chart so it's clear that Tesla is operating with safety drivers (in CA) / operators (in TX). The goal of this chart was to highlight the different services and where they're operating, and also what stage they're at without cluttering the actual chart too much. So hopefully this achieves that now.
Fair point. Tesla's operating with safety drivers and a waitlist, not driverless. Footnote 1 calls this out, but I should make it more visible. Thanks for the feedback.
We've now updated the chart to make the caveats around how the services operate clearer for each, right on the graphic. Thanks again!