Questioning the rosy predictions about the transition to autonomous electric vehicles

The Stanford economist Tony Seba and tech investor James Arbib just released a report entitled “Rethinking Transportation,” which makes an number of predictions about the impact that autonomous electric vehicles will have on the demand for vehicles and petroleum. Many of these predictions are based on faulty assumptions about human behavior and a misunderstanding of the auto supply chain.

The report predicts:

  • Private car ownership will drop 80% by 2030 in the US.
  • The number of passenger vehicles on American roads will go from 247 million in 2020 to 44 million in 2030.
  • Using electric ridesharing will be 4 to 10 times cheaper per mile than buying a new car by 2021 (and each family could save up to $5,600 per year, compared to purchasing and maintaining a traditional vehicle).
  • There will be a global decline in demand for petroleum, which will peak in 2020.

Seba and Arbib make a number of questionable assumptions to arrive at these rosy predictions. Autonomous electric cars will likely lead to more use of private vehicles, not less, for the following reasons:

  • It is more convenient to call an autonomous vehicle from a ride service like Uber or Lyft than use public transport, so there will probably be more cars on the road as autonomous cars replace buses and vans.
  • Autonomous driving mean that people can do other things while in the car (internet surfing, watch movies, work, rest, etc), so they won’t be as bothered by more traffic and longer commutes. Electric cars means that the air will be cleaner and there will be less road noise, plus the multimedia and internet services added to cars will make them more compelling environments, so people will be willing to spend more time in the car, not less.
  • Seba and Arbib assumes that cars will stay at the current 4% utilization rate for private use, but autonomous driving probably means that the percentage of time a vehicle spends in private utilization will increase, so there will be fewer cars available for renting out to the fleet than Seba and Arbib predict.
  • Most driving occurs during the peak hours during the morning and late afternoon when people are going to and from work. Faster internet that allow people to work remotely from home may reduce driving to a small degree during the peak hours, but most people who drive during those peak hours will still need a private vehicle. Most people who are willing to rent out their vehicles to a ride-sharing service will still need to own their own vehicles for their own transport during the peak hours, so demand for private vehicles will not drop as Seba and Arbib predict.
  • Cars will remain a status symbol and people will still want to own them. Look at the status of owning a car in Singapore, despite the fact that it can cost over $100,000 just to get a license to own a private car in Singapore for 10 years. If people were rational economic actors, Seba and Arbib may be correct in the long term about the  demand for private vehicles. Cars, however, convey social status and are packed with emotional meaning, such as feelings of freedom, independence, accomplishment in life and personal expression. People judge others based on their ownership of a car and its type, so the decision to own a private vehicle is more than a purely economic decision. It is unlikely that people who have grown accustomed to private vehicle ownership will be willing to shift their attitudes as quickly as Seba and Arbib predict.

Seba and Arbib may be right that the demand for private cars will decrease in some countries like the US and Germany which already have high ownership rates, but it is likely to be a gradual dropoff and not the kind of collapse that they foresee. In their predictions, they largely ignore the increasing demand for private vehicles in developing countries which represent 80% of the global population, which will more than make up for any decrease in demand among the 20% who live in the developed world. Almost all the auto companies are global companies, so they are not likely to go bankrupt if demand slackens in 20% of the world. I find it highly unlikely that the auto industry will be producing less than its current output of 90 million vehicles per year in 2030. In all likelihood, demand for vehicles will continue to grow at a rate of 4% per year, but even if all the factors that Seba and Arbib site come into play, it will probably lead to flat growth or a very gradual decline, rather than a dramatic collapse in global demand.

Seba and Arbib assume that the auto industry will function like most tech industries, with dramatic changes over short time periods, but there are number of reasons to believe that the auto industry will transition much slower than they predict to autonomous electrics. First of all, only Telsa, BYD and Nissan have laid the requisite groundwork to produce electric vehicles on a massive scale, and those three companies represent a very small percentage of the entire auto industry. We will need a 100 gigafactories producing batteries to make the full transition to electric vehicles and it will take at least two decades to ramp up all that battery production. There will have to be a massive increase in lithium, nickel, cobalt, copper and manganese mining worldwide, which will also take at least two decades to ramp up. Producing millions of electric vehicles per year will cause a huge spike in demand for these 5 metals and massive supply problems, which will  jack up the prices of batteries and will slow down the transition.

Remember that it can take years to open a new mine and there aren’t that many places in the world with high concentrations of nickel, cobalt and lithium. Currently the battery in a long range electric car needs roughly a thousand pounds of metals. By 2030, we might be able to get that down to 600 pounds of metals. Multiply 600 lbs by 100 million cars per year in 2030 and you start to see the scale of the metal supply problem. Even if we discover a new battery chemistry based only on common metals like iron and aluminum, it will still take decades to do the R&D and ramp up production.

Another thing that Seba and Arbib do not include in their predictions is how electric cars will lower the value of used gasoline cars. As autonomous electrics start to appear on the market, the resale value of gasoline cars will collapse, so it will be possible to pick up a used gasoline car for a couple thousand dollars, compared to buying an autonomous electric which costs $30K or more. People often assume that the price of autonomous electrics will keep dropping until they get to the $15K – $20K price range of cheap gasoline cars, but that may not be true, since the advanced sensors, the large amount of computing power, wiring and electronics, and the tremendous amount of resources in the batteries may prevent the price from dropping to the level of the current gasoline cars. Much of the growth in vehicles will be in 3rd World countries, where the elite market may become quickly saturated and cost becomes the biggest factor for the rest of society. People who want to save money may choose used gasoline cars, because the cost of buying a vehicle far outweighs the price of the fuel, even if electricity is significantly cheaper than gasoline and diesel.

Seba and Arbib are probably right that electric cars will decrease demand for petroleum, but this will also decrease gasoline and diesel prices, so there probably won’t be as large of a price gap between electricity and gasoline as they predict. Electric cars may be cheaper over the entire lifetime of a car, which will influence the buying decisions of fleet managers and transport companies, but their high upfront costs make that difference less important for the average consumer. Only if governments subsidize new electric vehicle purchases or guarantee the financing of electric car loans, can we expect consumers to buy new electric vehicles for economic reasons.

Most of the traditional auto companies want to protect their existing investments, so they will drag their feet as long as possible in making the transition to autonomous electrics. Some of them may go bankrupt because they are unwilling to change, but the chaos of bankruptcy proceedings will still slow down the transition. The big three auto companies, VW, GM and Toyota, that each produce 10 million cars a year, will take at least 15 years before they can retool to switch their production to majority electric, even if their management becomes fully convinced that it is necessary, but at this point they still aren’t convinced.

Even if the management of every auto company becomes convinced to invest all their available resources in transitioning to electric vehicles, I can’t see the world producing more than 2.5 million electric vehicles in 2020 and more than 15 million in 2025 due to global limits on battery production. By 2030, we might have enough battery factories and metal mining in operation so that the majority of the new cars in the world could be electric. Even if that happens, it will be another 2 decades to change most of global fleet to electric. The transition to autonomous vehicles could potentially happen faster, since it doesn’t face the same supply chain and dramatic retooling challenges, but it is unlikely.

Seba and Arbib are right to point to the future transition of the auto industry to autonomous electrics, but they dramatically underestimate the time to make this transition. Manufacturing automobiles is not like making silicon chips, cell phones, PCs or solar panels, and the auto industry can not make radical transitions on a dime like those tech industries. No auto company can call up Foxconn in China and ask them to produce 10 million cars and then slap their logo on top and ship them all around the world, like Apple does with iPhones. Seba and Arbib also misunderstand how consumers will respond to autonomous electrics, with increased personal driving, not less.

I say all of this with great sadness, because the world needs to transition as fast as possible to cleaner transport in order to confront climate change. I have never owned a car in my life and I have lived in Bloomington IN, Colorado Springs CO, Austin TX, Cuidad Juarez, Mexico and La Paz, Bolivia, so I know that it is possible in many parts of the world. Nonetheless, I also know the hassles in not owning a car and I suspect that the majority of North Americans will not be willing to give up their personal vehicles as Seba and Arbib predict. If it happens, the change will occur gradually, and be accompanied by a slow generational shift in attitudes. People who grew up and have used private cars all their lives are not going to change overnight. The social fabric and the urban landscape of North America is not going to change that fast to accommodate people without private cars. Employers are not going to switch that fast to the flexible working hours that would be necessary to eliminate the majority of the private vehicles on the road.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s