The effect of Uber’s surge pricing

[Editor’s note: This post is by Jonathan Hall, head of policy research, and Cory Kendrick, data scientist at Uber. It was published in September on the release of a case study looking at the real-world effects of Uber’s dynamic or “surge” pricing. We’re re-publishing it here.]

If you live in a city, you’ve probably experienced trying to find a cab on a busy weekend night, a holiday like New Year’s Eve, or during a surprise thunderstorm. It’s hard! Because when lots of people want a ride at the same time, you often have to wait a long time because all the available taxis are already busy.

Our goal at Uber is to ensure you can push a button and get a ride within minutes — even on the busiest nights of the year. And due to surge pricing, that’s almost always possible. Here’s how it works.

When demand for rides outstrips the supply of cars, surge pricing kicks in, increasing the price. You’ll automatically see a “surge” icon next to the products (uberX, UberBLACK, etc.) that are surging. If you still want a ride, Uber shows the surge multiplier and then asks for your consent to that higher price.

Surge pricing has two effects: people who can wait for a ride often decide to wait until the price falls; and drivers who are nearby go to that neighborhood to get the higher fares. As a result, the number of people wanting a ride and the number of available drivers come closer together, bringing wait times back down.

Together with Chris Nosko, a professor at The University of Chicago, we have been studying the effects of surge pricing. On New Year’s Eve last year, Uber experienced a technical glitch causing surge pricing in New York City to fail for 26 minutes. This created what we call in economics a “natural experiment” — when something varies, which you can then study after the fact.

Today we are releasing a case study of rider and driver behavior during the surge glitch, and on the night of a sold-out-concert at Madison Square Garden when surge worked as intended. This study is not exhaustive, but will form the basis of more comprehensive research in the future.

We found that, without surge pricing, Uber is not really Uber — you can’t push a button and get a ride in minutes:

• On the night of the concert, even though the number of people opening the Uber app experienced a 4x increase, the number of actual ride requests only rose slightly. In other words people decided not to request a ride. Meanwhile, 100% of ride requests were completed and ETAs were virtually unaffected.
• By comparison on New Year’s Eve, without surge, ride requests skyrocketed and only 25% of these requests were completed. ETAs also increased sharply. Without surge pricing, rider and driver behavior did not adapt to the increased interest in getting a ride.

These two real-world scenarios illustrate a bit of Economics 101: supply and demand adjust in response to price changes. On Uber, this means a ride is more likely than not just a few minutes away, at the simple touch of a button.