Tired of reading about the Uber-Travis-SoftBank drama? Try this working paper from John Horton at NYU Stern and two researchers at Uber Technologies Inc.
It argues that it’s pointless for Uber to raise fares, because the market will relatively quickly return to an equilibrium where drivers aren’t making any more per hour than they were before.
Here’s the logic: Imagine Uber normally charges $1 per minute and $1 per mile, and it increases that to $2 per minute and $2 per mile (these numbers are obviously made up). In the short term, drivers will earn a lot more, because, well, both rates have doubled. But over time, a couple things happen. One, because Uber suddenly lets them earn more, drivers work more hours. Two, because Uber is now more expensive, people order fewer Ubers.
When supply (driver labor) goes up but demand (trip requests) goes down, there’s less work to go around. Although drivers are getting paid more per trip, the overall time they work-the study calls this “utilization”-falls. After about eight weeks, their hourly earnings (“marginal revenue product”) return to what they were before the fare increase.
“We find that when Uber raises the base fare in a city, the driver hourly earnings rate rises immediately, but then begins to decline shortly thereafter”, the paper states. “The main reason is that driver utilization falls; drivers spend a smaller fraction of their working hours on trips with paying passengers when fares are higher.” (Uber has made the opposite argument to defend fare cuts, saying lower rates boost customer demand, allowing drivers to increase their utilization and hourly earnings.)
Changes in Uber’s base fare index don’t actually change marginal revenue product (MRP) in the long-run, where long-run is about eight weeks. (Horton, Uber)
Research papers published by Uber and its collaborators tend to talk about Uber like it’s a perfect free market, led by the invisible hand. This ignores that Uber has its own hands in the mix, texting drivers and offering them promotional payments to get on the road during busy periods, and hyper-targeting “upfront” prices to customers. (This paper takes place in a pre-upfront pricing universe, where what the rider paid was still the base for what the company and driver earned on each trip.)
The Independent Drivers Guild, a non-union group that represents ride-hailing drivers in New York, slammed the study. “The notion that Uber has no way to increase driver wages is absolutely ludicrous”, guild director Ryan Price said in a statement. “The company is pocketing higher fees than ever before, as high as double what the driver is paid for a trip, while drivers get pennies. If there is room for 220% Uber fees and projects like flying cars, there is room for a raise for drivers.”
Horton’s response to the drivers guild: “So here’s what’s wrong about how they’re thinking about it”, he said. “They’re not considering the fact that more drivers would drive more” if Uber raised wages. “If you basically said, alright, so now we’ll pay more-promotional payments, incentives, however you want to do it-as long as Uber runs an open platform, that’s going to get dissipated.”
Horton argues that there are things about Uber’s platform it controls completely, and things that are subject to constraints. What the customer gets charged for a ride, he says, “is basically 100% under their control, they decide.” What a driver makes from an hour of driving “is affected by the choices they make, but ultimately the decisions of other drivers are going to matter.” Uber has “a whole bunch of tools that shape what equilibrium emerges, but that doesn’t mean that every equilibrium is possible. It’s a constrained problem for them.”