Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

The economics of self-driving

This site may earn commission on affiliate links.
People on this forum have been generous enough to share their research on the technical aspects of self-driving with me. I thought I’d share this article I wrote, which presents my research on the economics of self-driving. I put way more hours into writing this article than I will ever be paid for it, and it is my pride and joy.

One of my most startling findings: the land value of all the parking in Los Angeles County is $1.8 trillion. With a ‘t’. That value is waiting to be unlocked by the post-parking world of autonomous ride-hailing. Self-driving will transform cities in a wonderful way, I think.

A hopeful prediction from RethinkX: self-driving cars will save Americans so much money currently spent on cars that it will be the equivalent of a 10% pay raise. I really think self-driving cars could be the beginning of a new era of progress and prosperity, where significant gains will be seen by everyone, including the bottom 20% of the income distribution.
 

I have some questions, if that is OK?

"There are over one billion unpaid drivers worldwide whose unpaid labour will be replaced by paid services, amounting to hundreds of billions or trillions in annual revenue for the companies that operate those services."

For me to pay a car-operating-company (and gain my unpaid-driver hours back) I will sell my car. So seems to me that I will be translating my current financing of a car into the hire of ride, so I see it as only shifting that money (from finance company to ride-hailing company) plus/minus the Finance and Ride-hailing markups.

"The resources allocated to manufacturing a car will be used more efficiently, spread across 5 to 10 times as many people and miles"

I get that; my car sits idle for 23 hours a day ... but a car doing 5x as many journeys/miles is going to wear out 5x as quickly, so apart from an initial drop in vehicle production when Autonomous becomes available, I think the replacement-rate will be similar to now.

" Over the long term, total miles traveled could increase between 50% and 3x. The total number of vehicles will shrink commensurately."

Why do you think people will travel more? If it is simply cost (assuming that Autonomous travel will be sufficiently cheaper) then won't that only applies to people who's travel is currently restricted by cost? However, I have no idea what percentage of the marketplace that is ...

"Parking will be eliminated, freeing up real estate in cities"

I don't get that. Certainly an autonomous car could drive itself out of the city to park, and then come back to pick me up (i.e. more cars in town supporting demand during rush hour, fewer during off-peak), but they may well need some parking in-town too. There is a cost-benefit equation in there somewhere for the cost of additional road-space and fuel to drive to an out-of-town car-park & recharge location. Obviously that works better in small towns where outskirts are only a mile or two away.

Maybe Musk-tunnels will whisk cars in/out of town?!
 
@WannabeOwner, I'll venture a couple of thoughts.

a car doing 5x as many journeys/miles is going to wear out 5x as quickly, so apart from an initial drop in vehicle production when Autonomous becomes available, I think the replacement-rate will be similar to now.
On a Tesla, for example, some items (like tires) will wear out at the same rate as a conventional car. However, the car itself is likely to last much longer (3x ?) than a conventional car.

Why do you think people will travel more?
Many people can't drive right now: underage, disabled, very old, for some examples. These folks make up a significant part of the population, and they are mostly new consumers for TaaS.

"Parking will be eliminated, freeing up real estate in cities"
Here's another way to think about that: in many cities, it's not a shortage of parking, but a misallocation of parking, either geographically or temporally, that drives the construction of new parking facilities. Clearly some situations (cities) are more efficient than others, but self-driving has the potential to completely change the relationship of car storage vs. other elements of real estate. It's hard to imagine, but consider the removal of cars from residential neighborhoods--or high rises--overnight. Residential construction could become *much* cheaper in Sunbelt cities built around the automobile. Cars could be stationed in common facilities (garages or lots that currently are empty overnight) and summoned when needed for transportation.
 
Thanks @Dr. J good food for thought

the car itself is likely to last much longer (3x ?) than a conventional car.

I expect that an electric car has many more miles "lifetime" than an ICE (and needs far less maintenance), but if it is doing 5x the miles (might well be 15x the miles, maybe more?), per day, because utilisation is increased dramatically, it ill wear out 5x quicker (i.e. wear out at probably a higher mileage than an average self-owned car achieves, but at a younger age). Thus it will be replaced "sooner". Maybe the overall fleet is dramatically smaller if everyone uses ride-hail?, but we still need enough vehicles for peak-time travel etc., and I can't quite get my head around the maths! but the replacement-rate for this higher-mileage, but smaller, fleet looks like being substantial - and perhaps not much less (once things settle down) that current car production - my guess is that my current car has a 15 year life (10-20K miles p.a. = 150K-300K miles), and a 24/7 ride-hail car only manages 5 years (100K miles p.a?. = 500K miles)

Many people can't drive right now: underage, disabled, very old, for some examples. These folks make up a significant part of the population, and they are mostly new consumers for TaaS.

I have no feel for how large that audience is, but it doesn't seem "big" to me - I am probably missing something?.

Clearly all that audience could, currently, take a taxi (I don't see an autonomous vehicle being more appropriate as a means of travel, although no doubt it will easier to get a nonstandard TaaS than a specialist Taxi at present - better density of such edge-condition vehicles - I, for one, won't be owning a People Carrier for the couple of journeys a month I use it on ... so those get shared within the local community).

I can see that some of them will not be able to afford the current Taxi price and if/when price drops they are then a buyer, But solely based on people that are impacted by that price-point doesn't feel to me to be a huge increase in (overall) journeys. For people who currently own a car they will still travel to work, and back, each day - they haven't got an opportunity to "travel more", and (to me) seems like they are the majority.

I can send my kid in a TaaS, instead of driving them there (and coming back again empty, and then going to fetch the little critter again later!), and whilst that's a gain of time for me, its actually a pair of (empty) drives less. Currently I would use a Taxi for that (if I didn't want to take them), or they would catch the bus, so doesn't seem like a change in journey-numbers.

I can see that more people may travel, more cheaply, if they ride-share, which TaaS is very likely to encourage, with a price saving, but that too doesn't equate to "more vehicles/journeys" of course.

self-driving has the potential to completely change the relationship of car storage vs. other elements of real estate

I'm interested in opinions about how this might change, in particular inner city car parks.

Let's say I use TaaS for my commute, and I'm the last person that vehicle has to carry during peak time. Where does that vehicle go to park? If it is "out of town" its going to clog up the roads on its journey (and use fuel, although maybe electricity will be abundant and that will be a non-issue). The alternative is that it just parks in one of the existing car parks. I definitely cannot get my head around "how many" vehicles might need that facility.

Perhaps everyone that "goes to the shops" for a few hours during the day will use TaaS such that it will be in equilibrium and not need to park at all, so I imagine this to be a commuter problem only - i.e; what to do with all the vehicles that are only needed for the commuting "load".

There being NO cars in residential neighbourhoods will be wonderful. We have lots of urban areas which were built long before cars were owned at the rate of "several per house". A large proportion of town areas have no off-road parking, so those streets will no longer be clogged with parked cars. We also have a problem that people convert their front garden to hard-standing, and then park there. That has caused problems with storm drains being under-sized for heavy rain due to the extra/changed rain run-off (i.e. no longer absorbed into soil). Maybe people will take up gardening again and the price of X-Box will collapse? :)

In UK I'm not sure what percentage of people actively use their garages (for cars), I think the majority are used to store "stuff". We don't have issues with extremes of temperature, or animals that want to eat their way into a car, here ... but for sure a building-opportunity for the conversion of garages no longer needed for cars.
 
I have no feel for how large that audience is, but it doesn't seem "big" to me

Well, I'm sure Texas is different from the UK in many ways ;), but, for example, our transit systems tend to suck, so many new TaaS customers will come from buses. The convenience of direct routes vs. indirect routes will be worth whatever incremental cost is incurred, and as you say, will likely be much cheaper than taxis (a service that also tends to suck here). My mom is disabled, so this kind of thing would be fabulous for her and for my nephew who has to take her to church and Wal-Mart.

We also have a problem that people convert their front garden to hard-standing, and then park there.
We have that problem here, too. While we are not as prone to flooding as Houston, it's a worry that is turning out to be extremely costly to address.
 
Last edited:
  • Like
Reactions: WannabeOwner
I'm interested in opinions about how this might change, in particular inner city car parks.
This issue is beyond my limited imagination. Somebody is going to make a lot of money on it, I have no doubt. I wouldn't want to be in the concrete business. Imagine the value of software that can control the efficient movement of fleets or the efficient usage of parking facilities. I imagine everything related to car storage gets smaller due to more efficient packing, usage, and in-and-out movements.
 
  • Like
Reactions: WannabeOwner