Self-Driving Cars Won’t Kill Demand For New Automobiles
Self-driving cars are coming, and they are coming quickly. The recent death of a Tesla driver in autonomous mode will barely delay the change. Despite that crash, self-driving cars are far safer than human-driven cars. But what does that mean for the car industry? I believe it will only trigger a small reduction in new car sales.
Fortune has a good article on Silicon Valley and Detroit racing to create “our driverless future.” It leads one to wonder, when cars are self-driving, why don’t we use Uber and Lyft and their successors instead of owning our own cars?
There’s good reason to doubt that auto-mobility will be completely Uberized, but suppose it is. We’ll still need a lot of cars to move us around. Those cars, though, will be used more intensively. After I arrive at my office, my car won’t need to stick around. Even if I’m meeting a colleague for coffee a mile away, going to lunch with a friend, and stopping by the gym on the way home, I will use a car—but not my own.
Cars being used more intensively, though, does not translate into less demand for cars.
The average car goes about 11,000 miles per year according to the Federal Highway Administration. Hard data on average life of a car is hard to come by, but 150,000 miles is a common estimate. That implies the average car lasts 13.6 years.
Now let’s say that service-provided self-driving cars are used much more intensively. A wild guess is a car is used four times as much, so that the annual miles driven goes from 11,000 to 44,000. Now the car must be replaced in 3.4 years. There can be fewer cars in use, but they will be replaced far more frequently. As a first approximation, annual car sales won’t change much.
We would be better off by having less capital tied up in cars. So take the value of the current car capital stock; multiply by three-fourths to capture the reduced capital tied up in cars; and multiply by the return on capital, as proxied by interest rates. That’s our annual savings from a smaller but more intensively used stock of cars
A more detailed approximation would account for dead-head miles. That is, after my car drops me off at my office, it may drive itself to pick up a student going to school. The trip from my office to the student’s home is an extra trip that never would have happened with self-owned cars.
On the other side of the coin, car services could offer more shared rides. If I’m headed downtown, the car service might tell me that with a two-block detour it can pick up someone else headed near my destination, cutting the cost of the ride in half. In effect, this is car-pooling without a schedule. Uber and Lyft have such a service now, but greater density of their activity would increase the number of instances of shared rides.
We might also save on time-related depreciation. Most of a car’s wear and tear and repair costs are mileage related, but some are time related, such as deterioration of rubber door seals and paint. With cars in service for fewer years, some costs will be a little lower.
Before we go overboard with Uber, though, go out to your car and look at what’s in it. We can ignore the old maps and the CDs that I listen to because I don’t want to fumble with electronic tunes while driving. I’ll still want my sunglasses. And my rain jacket in the back seat (because I live in Oregon). And all the little stuff in the center console: aspirin, nail clippers, notepad, breath mints. And all the stuff in the trunk, including things I’m going to drop off at Goodwill one of these days, umbrella, hat, sailing gloves, etc. Do I want to give up having these things handy?
When I head out of town on the weekend, I’ll hook my boat trailer to a van loaded up with sails, tool box, spare parts, and other stuff. Right now I can keep much of this stuff in the van; with a car service, I’ll schlepp it between my garage and the van every time. Maybe I should just own a van myself.
Some people love cars. Harvard Business professor Theodore Levitt is quoted as saying “People don’t want to buy a quarter-inch drill. They want a quarter-inch hole!” The good professor never visited the tool department at Home Depot, where do-it-yourselfers are drooling over 12-volt keyless-chuck variable-speed cordless drills. Similarly, cars are not just about getting from one place to another. I remember many years ago driving in a convertible sports car on a sunny spring day through the hills of Virginia with a beautiful lady by my side. Hard to have that experience with Uber or Lyft. Some people will want to own—and drive—their own cars.
Car sales will drop somewhat in the transition to self-driving cars, but the greater threat may be improved quality. With more precise manufacturing, cars are lasting longer. Tripping over 100,000 miles on the odometer used to be memorable; now plenty of cars are lasting 200,000 miles or more.
Finally, this article has been written from a first-world perspective. Rising incomes in third-world nations will trigger large additions to the automobile fleets of the world. The greatest challenge there will be roads to accommodate all the cars. With this stimulus, the global auto industry won’t shrink much, but it will certainly change.