Economists Do It With Models

Warning: “graphic” content…

Bookmark and Share
Peak Surge-Pricing Discussion, And An Open Letter To Uber…

November 6th, 2015 · 1 Comment

OH MY GOD YOU GUYS…based on my informal data collection, I really think that we’re approaching peak Uberbitching- I’ve been called upon multiple times in the past week or so to talk about this issue and why it gets people’s panties all in a twist, and I’m not even that popular. (It’s also funny because an economist is possibly the least-qualified person to explain why a scenario that approximates a free market would be bad.) Case in point:

(For the record, Nora has the best radio voice ever, and I am not too proud to admit that it does give me a bit of a hankering for Alec Baldwin’s Schweddy Balls.) I kid because I love- I am happy to talk to anyone about anything economics-y, I just might not always say what you want to hear. 🙂 I’m also sort of convinced that people are hearing about the Uber research done at Northeastern, googling to get more info, and finding me by accident. (If that’s true, hi!)

Overall, I HAVE THOUGHTS…more thoughts that can fit into sound bites, so I’ve tried to organize them here:

  • Surge pricing is just a somewhat more real-time version of dynamic pricing, which we see all the time nowadays with airlines, hotels, and even public transit when systems set different peak and off-peak rates.
  • With surge pricing, customers are gambling on prices, but, without surge pricing, customers are gambling on how long they will have to wait for a car. The second situation seems worse for people who are time constrained, and non-time-constrained people have the option to wait in order to avoid surge pricing, so it seems like gambling on price is the less bad option. (And no, you don’t have the option to gamble on neither, that’s not how any of this works.)
  • I’m not a monster- I do get why surge pricing annoys people. Most logically, the round-trip aspect of transportation likely has something to do with this- if you decide to go somewhere that requires an Uber to get back, it’s more than a little annoying that you can’t know how much your return trip is going to cost. People seem more or less okay with dynamic pricing more generally when they can see prices ahead of time and plan accordingly.
  • The case for surge pricing is strongest when the higher price both increases the number of cars and causes some riders to back off. In practice, however, surge pricing appears not to do much regarding the overall number of drivers (i.e. supply), likely because the real-time nature of surge pricing means that drivers can’t plan and adjust their schedules in response to price incentives.
  • Fun but often overlooked fact: The observation that most of the response to price surges happens on the passenger side implies that consumers don’t have Uber as their only option. (Or, as mentioned earlier, perhaps they’re just hanging out waiting for surges to go away.)
  • The fact the surge pricing mainly affects demand doesn’t mean that surge pricing doesn’t “work”- economists often talk about how prices serve to ration goods and services and, more optimistically, how markets work to get goods and services to people who value them the most. Using surge pricing to bring supply and demand in balances does, under certain assumptions, get cars to the people who value them the most, even when such practices don’t change how much of the good is available.
  • The above observation is based on the assumption that how much a person is willing to pay for something is a good proxy for how much that person values a good or service. In markets with a lot of income inequality, this may not be a valid assumption. Therefore, if people perceive high levels of income inequality, they may (rightly) not see the value, logic, or fairness in getting goods to those who are willing to pay the most. (I will, however, encourage you to put yourself in the shoes of someone who needs to get to the airport RIGHT NOW before dismissing the relationship between need and willingness to pay or the fairness of rationing via prices.)
  • I definitely hate surge pricing in sheep’s clothing more than I hate surge pricing- as a behavioral economist, I know full well what companies are doing when they offer a discount off of a normal price during slow periods rather than a surge price during high-demand periods. Similarly, please don’t brag about not doing surge pricing and then tell me that I can “boost” my ride to get a car to me faster by paying a fee, it just insults my intelligence.

I can ask for things I want on the Internet, right? I guess I always figured that that is what the Internet is here for. Well, that and screaming into the void. Uber and I aren’t friends on LinkedIn yet, so I’ll just leave this here for now:

Dear Uber,

As an economist, I am certainly not against the idea of surge pricing on principle. That said, unpredictable price changes make it really hard for both drivers and riders to make rational decisions and maximize their welfare with regard to your service. Therefore, I would really appreciate it if you would try one of the following, in order of priority:

1. Based on what I’ve come up with after thinking about this issue for way longer than a reasonable person should, it occurs to me that Uber should be able to get more drivers on the road using demand information rather than price increases. While it is true that workers tend to more more when wages are higher, it’s important to keep in mind that an Uber driver’s effective wage depends not only on the surge multiple but also on how quickly a driver can find passengers. In periods of high demand, it stands to reason that drivers should be able to find riders more quickly than during slower periods. This feature by itself actually creates higher effective wages during busier periods without the need for a surge multiple! As these drivers’ (not, as per legalese, but in spirit) employer, can’t you just provide information to drivers that signals times of high demand and remind them that it’s easier to make money when there are lots of available customers? You could even do this with more than zero seconds of notice in a lot of cases, since I’m guessing your data shows some pretty regular patterns in demand. It seems like drivers don’t respond much to the price surges anyway, so it’s likely the case that there is at least as much of an information/logistical problem as there is an incentive problem.

2. I’m guessing that, as a consumer, you wouldn’t want to plan to go out to dinner and have the restaurant refuse to tell you what its prices were until you’re about to sit down to eat, and you might still be annoyed even if you were cool with the restaurant changing prices to keep supply and demand in balance. In this spirit, won’t you please consider communicating an anticipated surge schedule to drivers and riders? You could still deviate from it when there are truly unexpected events (snowstorms, etc.), but this would go a long way towards helping market participants make price-based decisions. It would also likely mean that smaller price surges could keep supply and demand in balance since drivers and riders would be better able to change their behavior in response to price changes (i.e. could be more price elastic). I know that lower prices don’t necessarily sound good to you, but you’d be getting your 25 percent on more rides than you would with a larger price surge, so you could actually end up better off. If you wanted to be really cool, you could even develop a system to let people lock in a surge multiple- their return trip from that out-of-the-way location could still send them into bankruptcy, but at least they’d know ahead of time.

In conclusion, I really do believe that, in many circumstances, markets can be very powerful tools to create economic value, but I also believe that a well-functioning market would take into account the logistical realities of the environment it is operating in.


P.S. UberPool is awesome, but I might just think that because I really like being quoted a flat fee in advance.

And now we wait.

Update: We have some more information thanks to an economist who went the extra mile, literally…

Tags: Uncategorized

1 response so far ↓

Leave a Comment