Bike Sharing in China

The Flying Pigeon bicycle is the most popular vehicle model in human history, with more than a half a billion units deployed. Yet, bicycle use in China had been in steady decline since the 1990s. Indeed, the aspiration of most city-dwelling Chinese has been to own a car, which has contributed to clogged streets and air pollution in first-tier cities. Remarkably, bike sharing may spell the return of the bicycle in China.

Six months ago when I was last in China, few shared bicycles were on the roads and there were just a few promising signs of adoption. Today about 500,000 shared bikes are in use in Beijing, with similar penetration in other first-tier cities. At least five separate companies are vying for dominance. Users are crazy about the service. Traffic and subway lines are noticeably better. I’ve never seen adoption of a hardware-based service happen so quickly. What explains this remarkable early success and what does the Chinese experience imply for other cities?

@ktulrich on a Bluegogo shared bike in Beijing.

Some Background

As in Paris, New York, and other major global cities, Beijing (and some other first-tier cities) had installed kiosk-based bike share systems, in which the intelligence of the system is embodied in a fixed station, as shown below. The user identifies him or herself and checks out a bike, which is unlocked so it can be removed from its station. After completion of the trip, the user returns the bike to a kiosk at the destination. In the U.S. these systems cost about 5000 USD per bicycle to install and require long approval cycles with municipal governments, often three years or more. I am in investor in a rival approach by Social Bicycle in which the smarts and the lock are included in the bike. This allows the bike to be locked to any fixed object, although in most locations, the system includes “dumb” stations to which the bike must be returned (essentially a designated bike rack). This reduces the cost significantly, to about 1500 USD per bicycle.

Smart kiosk bike share systems, the norm in most of the world as of 2017.

This idea of a smart bike and a dumb station was taken to the next level by Mobike, a start-up that entered the Chinese market in 2016. (Mobike has raised more than 300 million USD in venture financing.) The first generation Mobike adopted several unusual design features, including an enclosed shaft drive, die-cast aluminum wheels, and a single-sided fork — none of which is essential to the basic business model, but which made the bikes quite distinctive on the street, and possibly reduced the risk of theft.

In 2016 the Mobike system was introduced in China. The lock simply immobilizes the rear wheel and a smartphone provides the user interface for locating, checking out, returning, locking, and payment. No fixed stations.

Bike Sharing 2.0 in China

Most observers I spoke with at the time of the Mobike launch were deeply skeptical that the bikes would survive on the street. I shared that skepticism when I first saw the bikes in Shanghai last year. I estimate these bikes, even in high volume, cost about 200 USD each, so significant losses would doom the model. But, remarkably, essentially no one stole the distinctive orange bikes. Furthermore, the average bike made about seven trips per day, exceeding expectations for capacity utilization. We were just wrong.

Competition quickly followed. Significant rivals include Ofo and Bluegogo. Ofo is bright yellow and Bluegogo is, well, blue. (Distinct color tones are quickly being grabbed by competitors, so I expect to see pink, white, and black models soon.) A typical streetscape is shown below, with various competitors’ bikes lined up along the sidewalk.

Typical streetscape for a bike-sharing city in China. Orange, yellow, and green bikes are from three different competitors.

Ofo pushed the model even further. The bike is so, so cheap. I bet it costs less than 30 USD per unit to make. There are no smarts of any kind. The user simply scans a QR code on the fender with the Ofo app and a combination code is revealed. The code is used to open a plain-old combination lock on the rear wheel. (At this writing, I’m still not sure how the system knows when the ride has been terminated — maybe there is a tiny bit of smarts in the system to send a signal that the lock has been deployed, but I can’t see where it would be. Or maybe it always charges a flat fee of 1 RMB.) Bike sharing in China is almost free. The charge is about 1 RMB per hour (with various fixed+variable pricing models for different services). That’s about 0.15 USD per hour. Most trips are of course less than an hour, so a trip costs 0.15 USD for most users. By comparison, the subway in Beijing, cheap by global standards, is a minimum of 3 RMB (0.50 USD) per trip.

Those I spoke to in Beijing, including venture capitalists who financed these businesses, claim there is no government subsidy (other than allowing these bikes to be parked on public sidewalks and the maintenance and provision of bike lanes on most major thoroughfares). But, I think the economics actually pencil out. The bike earns about 7 RMB per day (1 USD) so only needs to be used 30 days to cover the manufacturing cost. Of course there are other costs, including night-time redeployment and repairs, but labor is still relatively inexpensive in Beijing (about 3 USD per hour) and so these costs can be covered. With bikes this cheap, there is no real reason to invest heavily in GPS tracking, fancy locks, and parking stations.

Mobike still makes a nicer bike, but their current version (photo below) is much more conventional, and presumably much less expensive, than the first model. Notably, Mobike has gone to a sold urethane tire with lateral holes to provide some additional cushioning. This really reduces the most common maintenance problem — flat tires. I found the tires to be just fine for the short rides I made.

Mobike model as of 2017 — note solid urethane tires with lateral holes, conventional chain drive, and conventional fork.

We can observe a lot of parallel experimentation by rival companies. The “green” company (whose name I don’t know — possibly Haoqi based on the好骑 I see on the back fender in the photo below). This bike looks more like the Social Bicycle design, which includes a solar charging feature and a full mobile computer inside. Likely the Chinese market will converge soon on a dominant design and just two or three companies will survive.

More feature-laden options in the “green” bike (好骑?)

What explains the amazing rate of adoption of bike share in China?

Here is what I think explains the bike share explosion in China.

Bike share addresses an acute and deep pain point. At 5pm in Beijing, prior to bike sharing, the line for the subway in the central business district snaked out onto the sidewalk requiring more than 30 minutes just to get to the entrance point. Taking an “Uber” (actually a “Didi” in China) involved sitting in a traffic jam. A bike is simply the fastest and cheapest way to get from A to B. It is 3x less expensive than the subway and may cut the travel time by 5x or more for distances of up to 3–5 km (2–3 miles). The solution is a “pain killer” not a “vitamin pill.”

Chinese cities are extremely dense allowing high capacity utilization and high availability. Within a city block there may be 100,000 people. That means that every street corner is effectively a bike share hub that can have 20–50 bikes available at any time. This is not going to work in Topeka, nor probably even in Portland. It might work in New York.

The government, while not directly subsidizing bike share, allows it. It is hard to imagine the City of New York allowing laissez faire bike sharing on its streets and sidewalks. One would think that the communist state would be even more controlling, but in this case, bike share is so popular and solves such critical problems (traffic, subway congestion, air quality) that the government allows unfettered competition in public spaces.

Labor is inexpensive enough to allow nightly redistribution and maintenance of bicycles. I have not modeled how important this is, but my gut feel is that the economics do depend on highly affordable maintenance operations.

The base rate of crime, including petty theft, is lower in China than in many other parts of the world. This is partly the result of culture and partly the result of enforcement of law and order. Still, given the ubiquity of the bikes, their distinctive appearance, and the low cost of legally using them, what is the incentive to steal one? Theft might prove not to be a problem in the U.S. either.

What About the Usual Objections to Riding Bikes?

I’ve been a bike commuter my whole life. Most people raise three objections to riding a bike for transportation: weather, sweating, and safety. Beijing is a northern city. It’s cold in the winter, but a lot of snow on the streets is rare. Think of it as similar to New York City in that respect. Beijingers have happily ridden motorbikes for decades. They dress warmly and don’t typically spend more than 15 minutes on any one trip. With a compelling enough benefit proposition, people will manage the weather. (I sometimes argue that people spend 10,000 USD to have cold wind in their faces — ski vacations — and that they could get that for free by riding their bikes in winter.) Sweating is minimized in China by riding very slowly. In the US we practically race our bikes! I’m trying to learn to ride more slowly. Safety is a big concern. I don’t think we have good accident statistics on bikes in China. I bet the rates of injury are quite high, but don’t have evidence one way or the other.

Challenges in China

The three main emerging challenges I see for bike sharing in China are safety, air quality, and the aesthetics of the streetscape. Safety concerns are obvious — not sure the solution. Maybe we’ll see adoption of stylish helmets that attach to a backpack for storage or something. Most of the air quality challenges in China are due to industrial particulate emissions not automobiles (e.g., coal-fired power plants), so the air quality concern relates primarily to physical exercise while breathing polluted air. In the long run, I believe China’s air quality will improve substantially, but I bet it takes 20 years. I have not researched the efficacy of the use of masks, which is widespread in China. Perhaps there is an opportunity here for a nice helmet with integrated filtration mask. In my opinion, the streetscape aesthetics are degrading with the use of bike sharing. See the photo below for a typical street corner. I think the municipalities could address this by carving out some parking areas. China does infrastructure really really well, so I can easily imagine this happening.

Typical street corner with unfettered bike sharing.

Some Practical Notes for Visitors to China

Most visitors to China will not be able to use the bike share system unless they have WeChat payments or Alipay set up on their mobile phones, and some apps require a Chinese ID card. That’s not so easy for most foreigners of course. If you want to try it, get a Chinese friend to unlock a bike for you. It costs them 1 RMB, so no big deal. Most systems do not allow one user to unlock two bikes, but most Chinese using bike share will have signed up for multiple services, so can unlock one Mobike and one Ofo bike, for instance, if you want to ride together.

Riding a two-wheeler in China appears very dangerous. (No one wears a helmet — I’m going to bring my own on my next trip.) The danger appears primarily because motorbikes share the bike lanes and even the sidewalks (see bike path in Chengdu below), and they often travel in the wrong direction. These motorbikes are all electric in Beijing and Shanghai, so essentially silent. Very disconcerting. My strategy is to ride very slowly, as most Chinese do, with the flow of a larger group of cyclists, and just yield to the cars and motorbikes doing their thing. Be careful out there!

Chengdu bike path, including motor bikes.

Prospects for the U.S.

The boom in station-less bike sharing in China has not gone unnoticed in the U.S. The Chinese company Bluegogo has entered the San Francisco market directly. Indeed, I first saw a Bluegogo bike in San Francisco a few days ago. The bike is slightly different from its Chinese counterpart, notably including a multi-speed transmission, presumably to accommodate the hills of the city.

Bluegogo bike spotted in San Francisco.

And Lime Bike recently closed $12 million in venture financing to bring the Chinese model to the U.S. Bluegogo’s original strategy was to just place the bikes on the streets with no coordination with the city. The San Francisco Municipal Transportation Agency (SFMTA) quickly noticed and made an inquiry of the company as to its intentions. The city then introduced new regulations for station-less bike-share systems. The regulation seems sensible, and SFMTA appears to be interested in stimulating better bike-sharing options, and so my prediction is that station-less systems will be allowed soon in many major U.S. cities.

Significant uncertainty clouds the prospects for the new entrants. Will the lower density of U.S. cities allow for sufficient availability of bikes? Will higher crime rates in the U.S. result in unsustainable losses? Will labor costs allow companies to preserve the attractively low prices offered in China? Will municipalities support effective and attractive parking solutions? Only time will tell.