Uber still doesn’t look like the next Facebook
Uber has been in the news a lot lately, and most of it hasn’t been pretty. Allegations of a sexist workplace culture, a high-profile legal battle with Alphabet Inc. (Google) over self-driving car technology, reports of attempts to skirt local laws, an anti-Uber Twitter campaign, and an exodus of top talent have put the ride-hailing giant on the back foot. But so far, these problems don’t seem to have put a huge dent in Uber’s market share—as the website TXN found in March, even after many of these problems hit the news Uber was still providing almost four times as many rides as its biggest rival, Lyft Inc.
Negative publicity and management squabbles are not helpful, but in the end, it will probably be the strength of the ride-hailing business model that determines if Uber—and rivals such as Lyft—live or die. And it’s here that Uber and its investors should be worried. Basic economics shows some cracks in Uber’s model. Those weaknesses might not be catastrophic, but they probably deserve more attention than they’re getting.
A company’s value depends on how much money it can make. Competition limits profitability, so to protect their margins, companies need what Warren Buffett calls a “moat”—a source of market power that provides an enduring advantage over new entrants to the market. The moat can be a valuable brand name, such as Gucci or Apple. It can be technological superiority or government-granted patents. It can be economies of scale, as with the auto industry—it’s hard to get the money to start a company big enough to compete with Toyota Motor Corp. and Ford Motor Co.
But the most powerful kind of moat is a network effect. I’m on Facebook because my friends are on it, and the same is true for them—we can’t easily switch to a new social network, since no one is there yet. When the value of being in the network increases with the size of the network, it’s a powerful source of market power. Technology companies, which take advantage of the connectivity of the internet, tend to have strong network effects, often resulting in winner-take-all markets.
Uber undoubtedly has a network effect. Each rider has an incentive to use the hailing app that has the most drivers on nearby roads, since that will minimize wait times. And each driver wants to use the app that has the most customers ordering rides, because that will minimize the time to find a fare. This is known as a thick market. Imagine a ride-hailing app with only three drivers and 100 users—it would usually take a long time to order a ride, because chances are that the car and rider would be far apart. But with 300 drivers and 10,000 users, wait times will tend to be low. Having a larger network is valuable.
But there are reasons to believe that this network effect may not be overwhelmingly strong in Uber’s case. First, as my colleague Justin Fox has pointed out, dominating one city doesn’t help that much in dominating another city, since only travellers order rides in both cities. Second, most Uber drivers also drive for Lyft or other Uber competitors, and many users have multiple hailing apps on their phones. The ease of switching between two apps means that Uber is unlikely to ever be able to force Lyft and other rivals out of the market.
This may be a big reason that Uber has failed to achieve profitability. In a long series of blog posts, consultant Hubert Horan documents Uber’s enormous financial losses, and provides some evidence to support the theory that these are due to operating losses, rather than investments intended to help the company grow. Most tellingly, Uber’s losses haven’t improved much as the company has grown, meaning that unlike tech companies such as Facebook Inc. and Amazon.com Inc. with strong global network effects, Uber and other ride-hailing services may not be able to grow their way to profitability.
One thing Uber could do is to raise prices. If Uber and Lyft are locked in a destructive and ultimately futile price war, they could potentially improve their bottom lines by simply ending that war and accepting that neither will ever be a monopoly. But that might choke off demand for their services, as people decide to switch to public transportation or drive their own cars. Some economists have estimated that demand for Uber’s services is very inelastic—that is, customers would be willing to pay higher prices. But as Horan points out, the study focused on short-term demand by frequent Uber users—in the long-term and across the whole economy, demand is likely to be much more sensitive to higher prices.
Another thing Uber could do is use self-driving cars, as it is currently attempting to do. But as asset manager Hamish Douglass has noted, this could make matters even worse if it succeeds. A human driver can fairly easily switch back and forth between two ride-hailing apps, but an automated vehicle can effortlessly manage a hundred at once. That immediately knocks out one side of Uber’s network effect, since every app will be able to hail every car. In addition, self-driving cars will take away much of the rationale for using ride-hailing apps in the first place, since people’s own cars will start to function as their own private taxi cabs—the stress and time of driving yourself around and finding parking will be gone, and your car will even be able to drive you home after you’ve had one too many.
So it may be that investors, and much of the business press, have overestimated the profit-earning capability of companies like Uber and Lyft. They or something like them will undoubtedly survive, but they may be more of a low-margin business than their investors have hoped. Bloomberg