Friday, 27 May 2016

Uber says they're not price discriminating - should we believe them?

James from my ECON100 class pointed me to this article about Uber, surge pricing and price discrimination:
The car-hailing service Uber can detect when a user’s smartphone is low on battery, and therefore willing to pay more to book a ride.  
Uber, which has faced the ire of London’s tax drivers since launching in the capital in 2012, can tell when its app is preparing to go into power-saving mode, although the firm says it does not use this information to pump up the price.
Keith Chen, head of economic research at Uber, told NPR that users are willing to accept a “surge price” up to 9.9 times the normal rate, particularly if their phone is about to die.
“One of the strongest predictors of whether or not you’re going to be sensitive to surge… is how much battery you have left on your cellphone,” he said.
“We absolutely don’t use that to push you a higher surge price, but it’s an interesting psychological fact of human behaviour.”
So, Uber has discovered that people are more willing to pay a higher surge price when their phone battery is about to die (see my previous post on surge pricing). This suggests that Uber passengers have less elastic demand when their phone is running low on charge. This makes sense for a couple of reasons, both relating to time horizons. First, if you need to get home (or to work, or to a meeting, etc.) and your phone is running low on battery, then you need to get a ride soon. If you wait too long, your phone might run flat and then you might find it more difficult to find a ride home (you'd need to resort to hailing a cab on the street, using public transport, or walking). Second, we don't want to be without our phones for too long, lest we miss Kim Kardashian's latest tweet. So, we want to get somewhere where we can recharge our phones, and fast. We know that when customers have short time horizons, their demand is relatively less elastic. And when demand is relatively less elastic, firms can mark up the price higher and the customer will still be willing to pay.

So, should we believe that Uber is not price discriminating? Price discrimination increases profits when firms can do it effectively. This only requires three conditions to be met:
  1. Different groups of customers (a group could be made up of one individual) who have different price elasticities of demand (different sensitivity to price changes);
  2. You need to be able to deduce which customers belong to which groups (so that they get charged the correct price); and
  3. No transfers between the groups (since you don't want the low-price group re-selling to the high-price group).
The first condition is clearly met, and presumably Uber's app knows when the phone is low battery (it's probably buried in the terms and conditions for the app, which almost no one reads). Since customers don't know the battery status of other Uber customers, then the third condition is likely to be met too. So, if Uber isn't price discriminating on the basis of battery level, they are leaving potential profits on the table. Uber shareholders probably wouldn't be too happy to learn this. So, I think it's hard to believe that Uber don't take a lot of information about their passengers (including potentially the remaining battery life of their phone) into account at least at some level - perhaps they are not price discriminating via the surge price (i.e. the multiple by which they increase prices), but via the underlying base price?

As an interesting side-note, later in the article we find out a bit more about the price elasticity of demand of Uber customers when surge pricing is new, and then later:
When Uber first introduces surge pricing to a city, even a small jump in price to 1.2 times the normal rate is enough to discourage 27pc of potential customers from booking.
However, cities with more established Uber services see a smaller 7pc reduction when this price rise kicks in, showing customers get used to the idea, Chen said. 
This suggests there is relatively elastic demand when Uber first introduces surge pricing to a city (since the percentage change in quantity (27%) is greater than the percentage change in price (20%)). But there is relatively inelastic demand once customers are more used to it (since the percentage change in quantity (7%0 is less than the percentage change in price (still 20%)). This suggests that people are initially resistant to temporary increases in price (through surge pricing), but once they realise the benefits of it, that initial resistance fades.

I also found this bit interesting:
Mr Chen also noted the quirks of psychology that make passengers more likely to ride when the price is 2.1 times normal, rather than 2.0 times, which chimes in their consciousness as being twice as expensive.
“Whereas if you say your trip is going to be 2.1 times more than it normally was, wow, there must be some smart algorithm in the background here at work, it doesn’t seem as unfair,” he said.
I've seen other research that supports this (but off the top of my head I can't recall where) - it showed that people are more likely to be confident in a forecast share price of, say, $2.02 per share, than a forecast of $2.00. People believe there must be some sophisticated underlying model for a forecast of $2.02, whereas $2.00 seems more like a guess. Aren't cognitive biases wonderful?


  1. I feel much better if I have a battery pack with me

  2. I feel much better if I have a battery pack with me