Electric Cars, Sooner or Later – Bloomberg View – Bloomberg View

Posted: Thursday, July 23, 2015

It’s always interesting to see debates between technologists and economists. Technologists are inclined to think about new possibilities, while economists are trained to think about tradeoffs between existing possibilities. This makes for a very compelling intellectual dynamic. One recent example of such a debate was between Y Combinator partner Geoff Ralston and George Mason economist Alex Tabarrok. 

Ralston says that when electric vehicles become cheap enough to be a real alternative to gas-powered cars, the network of gas stations will collapse in a stupendous “tipping point“:

Gas stations are not massively profitable businesses. When 10% of the vehicles on the road are electric many [gas stations] will go out of business. This will immediately make driving a gasoline powered car more inconvenient. When that happens even more gasoline car owners will be convinced to switch and so on. Rapidly a tipping point will be reached….Then, there will be a rush to electric cars[.]

Tabarrok doesn’t think this will happen, for two reasons. First, he notes that a modest decline in the number of gas stations will only bring us back to the level of a few years ago, when driving was (presumably) not that inconvenient:

 Why is [Ralston] wrong? Consider that in 2009 there were 246 million motor vehicles registered in the United States. A 10% reduction would be 221 million vehicles but that is how many vehicles there were in 2000. Was driving an automobile so much more inconvenient in 2000 than it was in 2009? No. Even a 25% fall in gas vehicles would bring us back only to the number of vehicles circa 1998.

Tabarrok is probably right — the tipping point that Ralston posits is far in our future, since there are so many gas-powered cars and so many gas stations in existence. But his argument overlooks one important factor: substitution. 

In 2000, there was no such thing as an affordable electric vehicle. If there were places you couldn’t find a gas station, it was a hassle cost, and the only alternative to paying that hassle cost was not to drive. But in the future Ralston describes, there will be an alternative — electric vehicles. If electric cars are cheap enough, then the hassle cost of finding gas stations would induce some people to switch vehicle types rather than simply drive less, triggering the cascade that Ralston predicts. 

Of course, this depends crucially on how electric cars are charged. As long as electric cars have limited range, they will need charging stations. And as long as charging stations are much less common than gas stations, that pushes the tipping point very far into the future. 

Tabarrok’s second argument against an electric-car tipping point makes less sense to me. He says that a reduction in the number of gas stations will cause an increase in the demand for gas stations, which will then increase in the number of gas stations:

More fundamentally, [Ralston’s] argument goes wrong by not thinking through the incentives. Suppose that gasoline stations did become so uncommon that finding one was inconvenient. What will happen? More gasoline stations will be built!…[G]asoline stations are relatively simple, small businesses that easily expand or contract in response to the pennies on a gallon that people are willing to pay for convenience.

This kind of thought experiment is a lot easier if you just think about supply and demand. Tabarrok seems to be claiming that the supply of gas stations is very price elastic, meaning that the number of gas stations will decrease a lot in response to a fall in the demand for gasoline. Here’s how that might work: The availability of an electric-vehicle alternative will reduce demand for gasoline, since a few first movers will switch to electric cars. That will cause a slight reduction in the price of gasoline. If Tabarrok is right, the number of gasoline stations will fall a lot. And since gasoline cars are a complement to gas stations — that is, they go together — that means purchases of gas-powered cars will also drop a lot. 

In other words, it seems like Tabarrok is mostly agreeing with Ralston on this second point. 

Ralston, however, is adding a wrinkle that isn’t normally considered in simple supply-and-demand analysis: network effects. Network effects are when the number of people in a network increases the value of being in the network. Think of social networks as the perfect example. Everyone uses Facebook because all their friends use Facebook. If someone created a new, slightly better version of Facebook, very few people would switch, because none of their friends would be on the new site. Thus Facebook maintains what amounts to a natural monopoly. 

Cars and filling stations are the same way. The more gasoline cars are on the road, the more demand there is for gas stations. The more gas stations are along the road, the more convenient it is to own a gas-powered car. The existence of this vast infrastructure — and the cost of building an alternative infrastructure — is the biggest reason electric cars will find it hard to break into the market. 

The one thing that might make a big difference is if electric cars get really long ranges and can be driven all day at highway speeds without stopping. In that case, electric cars won’t even need charging stations and can bypass the hurdle of creating the network effect. At that point, the network effect between gas-powered cars and filling stations would start to work against gasoline, just as Ralston predicts. 

So as I see it, the near-term future of the electric car depends crucially on battery breakthroughs that allow very long ranges. As to how close we are to those breakthroughs … you would have to ask a technologist.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author on this story:
Noah Smith at nsmith150@bloomberg.net

To contact the editor on this story:
James Greiff at jgreiff@bloomberg.net


Write a Reply or Comment:

Your email address will not be published.*