Last year, China accounted for nearly half of all electric vehicles (EVs) sold worldwide. China has offered generous subsidies for EVs and has invested heavily in charging infrastructure. Those policies would seem to explain the popularity of EVs in China, but lately those subsidies have been phasing out. Many large cities have moved to a different approach: giving preferential treatment to EVs in the allocation of new license plates. In those cities, consumers have to wait several years to buy new gasoline-powered vehicles, but they can often buy EVs much more quickly. A new paper by PhD candidate Yujie Lin shows that this approach is the most important policy driving EV sales in Beijing and Shanghai. Now, one might expect that this type of policy would be costly, but in this case the benefits of lower pollution outweigh the costs.

To explain what’s going on in these cities, I need to describe the license plate allocation systems. As I’ve written previously, rising income has caused vehicle ownership in China to take off; China has by far the largest market for new cars in the world. To reduce massive congestion and pollution problems that this car ownership creates, eight cities use lotteries or auctions to restrict ownership. Depending on the city, a household wishing to add a car must enter a lottery or auction for the license plate (a household replacing a car doesn’t need to enter). These systems impose costs on people by making them wait until they win the lottery or by having to pay the auction price. In Beijing, the average entrant may have to wait 5 years to win the lottery, and in Shanghai the auction price can be a third the price of the car (the auction winners pay that amount for the license plate, plus the price of the car).

The license plate lotteries may be more equitable than auctions, since winners are chosen randomly rather than based on willingness to pay (which is correlated with income). But auctions are more economically efficient for the same reason—with a lottery, some people who really need cars may be unlucky and have to wait. Colleagues and I have studied Beijing’s lottery, which has substantially reduced vehicle ownership, driving, and pollution. (Incidentally, the lottery has also reduced fertility, as the inability to own a car appears to cause households to delay having a baby.)

Many of the same cities give preferential treatment to EVs to help achieve China’s EV goals and reduce pollution. For example, Beijing has a separate lottery for EVs, and it’s much easier to win a license plate in the EV lottery than in the other one—hence, the wait time for getting a new EV is much shorter than the wait time for a gasoline-powered vehicle. In other words, Beijing and other cities are increasing the appeal of EVs by reducing wait times for a scarce resource—that is, a license plate. Shorter wait times cause some consumers to buy EVs who would have bought a gasoline-powered vehicle.

Yujie’s paper asks two questions: how much has the preferential EV system boosted EV sales in Beijing and Shanghai, and at what cost to consumers? To answer these questions, she compares actual EV sales and consumer well-being in those cities against what would have happened if the cities had a single system for all vehicles rather than a separate system for EVs. Consequently, she needs to consider how changes in wait times affect buying decisions. In other words, with the dual system (one system for gasoline and the other for EVs), the shorter wait time for EVs causes many people to buy an EV. If the dual system is replaced by a single system, how many people would change their mind and buy a gasoline-powered vehicle instead? The answer depends on whether they prefer gasoline-powered vehicles over EVs and how much they mind waiting.

Yujie answers these questions by comparing Beijing and Shanghai with two other cities that have similar demographics and vehicle purchasing patterns prior to the lotteries and auction—Chongqing and Suzhou. She estimates consumer preferences for gasoline-powered vehicles and EVs in those cities and estimates how much people don’t like waiting by comparing EV market shares among Beijing, Shanghai, and the two control cities. In other words, if EV market shares are higher in Beijing and Shanghai than the other cities, she concludes that people really don’t like waiting to have to buy a gasoline-powered vehicle.

Yujie finds that the vehicle ownership restrictions impose considerable waiting costs on consumers in Beijing and Shanghai, which equal roughly 10 percent of the price of the car. The preferential treatment for EVs explains half to two-thirds of EV sales in those cities. In other words, if Beijing and Shanghai replaced the dual system with a single system for all vehicles, EV sales would fall dramatically.

Interestingly, the welfare costs of the dual systems are fairly modest. The dual systems cause many consumers to buy EVs who would otherwise prefer gasoline-powered vehicles. That represents a cost because consumers aren’t getting the vehicles they prefer. But the general public benefits from the lower pollution, and those benefits outweigh consumer losses, so that on balance the dual system makes society better off. Of course, other policies like a carbon tax may be more economically efficient, but Yujie finds that this alternative approach is definitely better than doing nothing. That’s an important finding since this approach is such a big part of the reason for the success of EVs in China.