Dynamic pricing and fairness – Uber in Sydney

The car-sharing service Uber uses dynamic pricing (also called surge pricing): prices rise in response to local excess demand (where the demand for rides at the current price exceeds the supply of cars), and fall in response to local excess supply (where there are more cars available at the current price than there is demand for them). Uber’s sharp increase in prices in Sydney, where there was a sudden significant excess demand for cars as the result of a downtown hostage taking, has sparked a debate on the morality of dynamic pricing. It’s not an arts example, but since so many arts organizations are experimenting with dynamic pricing (my caveats are expressed here), I think it is worth looking into.

As a preliminary thought, notice that we see prices in competitive markets responding to demand and supply all the time: look at the variations at the grocery store, at the gasoline pump, in the markets for real estate and apartment rentals. If prices didn’t change, we would be constantly seeing severe shortages of some items and overstocks of others. But we don’t see that. And a good thing too. Rising prices in the face of excess demand encourage buyers to look to potential substitutes (as the price of beef rises people look to recipes for chicken), and at the same time encourage firms to increase their supply of the good, until the excess demand is eliminated. Further, the scarce good goes to those people who value it most highly: if beef prices rise relative to chicken, beef will continue to be bought by those with the strongest preference for it, while those who are somewhat indifferent to beef are the ones who switch to buying more chicken.

If Uber uses dynamic pricing, we get the same result, albeit in a situation where one company is doing the price adjustment. In the face of excess demand for rides, the increase in price draws more cars into service, and also encourages people to look for substitute means of travel. The ones who ride with Uber even at the high price would be the ones who see no viable substitute. If prices don’t rise when there is excess demand, we have a shortage, and the people who catch a ride are the people lucky enough to book a ride first. These are not necessarily the people who value the ride the most.

I don’t think anything so far is very controversial, it’s textbook stuff. And it is hard to see any moral objection to it.

Until we get to the situation of a crisis. If the demand for rides has sharply risen because of a crisis situation, as happened in Sydney, does surge pricing become immoral?

The Washington Post reported:

After an armed gunman took patrons of a Sydney chocolate shop hostage during rush hour Monday morning, the transportation company Uber quadrupled its fares for panic-stricken customers fleeing the central business district.

It charged a minimum of $100 ($82 U.S. dollars) to escape.

As the Sydney Opera House and other buildings were evacuated and images flashed on TV screens of hostages holding a flag bearing the message “There is no God but Allah” and “Mohammed is the messenger of God,” Uber tweeted it was raising prices.

The company has defended what it calls “surge-pricing” as necessary to encourage more drivers to pick up passengers when demand is high.

Within an hour, Uber backtracked after the media publicized customer complaints of price-gouging.

Olivia Nuzzi weighs in with a defence of Uber’s price increase:

The fact that Uber allowed surge pricing during a hostage crisis may lead you to believe that the company doesn’t care about you, and you would be correct. But Uber does not have a responsibility to care about you. Uber is not a government entity, and it is not beholden to the general carless public during an unwelcome drizzle of rain or even a time of great distress. …

The premise of the program is simple supply-and-demand: when demand for cars increases and supply decreases, Uber’s algorithm inflates the fee for rides accordingly, which the company claims encourages more drivers to work, which puts more cars on the road when people are requesting them most. 

Surge pricing happens during rush hour and when it rains or snows; it happens on holidays like Halloween and New Year’s Eve; and it has happened during states of emergency like Hurricane Sandy in 2012, after which the New York Attorney General stood Uber down and made them agree not to hike up fares during natural disasters––and he was as wrong as any of Uber’s customers who complain about their inflated fares. …

Uber’s surges are not price gouging, as some have erroneously claimed. Uber––which is actually not the only method of transportation on Earth, despite what it may seem like––warns passengers about the surge before it allows them to order a car, and if the surge is over two times the normal rate, the app forces users to type it in, just to make sure they really understand what they are getting themselves into.

‘Price gouging’ is a meaningless term, so I don’t see much point in getting into debate as to when a price increase is gouging or not. But that term aside, if a price increase during rush hour is acceptable, why not a price increase when disaster strikes?

At Slate, Alison Griswold writes:

Is it wrong to raise prices during an emergency? Here’s what Uber has said in the past, and repeated in Sydney on Monday: Surge pricing in times of need is important—even beneficial—because it encourages drivers who might otherwise stay safely at home to hit the streets. This is a fair point. If there’s any time that people should want drivers to be incentivized to pick up passengers, it’s during a disaster or emergency. Raising fares helps make that happen. Of course, the flip side is that as fares goes up, Uber becomes less and less reliable for anyone who can’t afford to drop $100 on a single cab ride. Surge pricing doesn’t make Uber more accessible to everyone during an emergency—just the people who can pay for it.

I don’t buy it (sorry). Rides are bought by those consumers willing to pay the most for them, yes, but that is true of every good and service in the market, from sirloin steaks to tickets to the Chicago Symphony. Without a price increase, the rides would be obtained by … whom? Essentially, those who got lucky. But that doesn’t strike me as a morally superior way to decide who gets a car, and it doesn’t make a difference what was the cause of the price increase.

At Harvard Business Review, Tim Sullivan looks to social norms:

Prices do a very good job of both aggregating preferences and giving weight to them. In their absence, it can be really hard to allocate goods, which is to say that not using the price mechanism is less efficient than using it. Just ask anyone who lived in a centrally planned economy. It’s also one of the reasons that the market can appear to be so much more efficient than our offices. The market uses prices, while our organizations tend to use command-and-control mechanisms to deal with situations often too complicated for the market to handle. (Shameless plug.) Uber is simply increasing the efficiency of the overall system, and those who don’t see and understand that should learn to.

But I’m pretty well educated about economics, and I still feel a moral twinge about the idea of surge pricing. So what’s going on?

First, I think it has something to do with how people react to prices. Recall the experiment at the Israeli day care [PDF] made famous by Freakonomics. Researchers found that when daycares charged a late fee, $3 for every ten minutes, more parents were willing to be late. What was going on? Once parents could put a price on their lateness – rather than feeling like they were inconveniencing the caregivers – they were willing to pay out of pocket. The moral compunction to be on time was stronger than the late fee.

In the case of the Sydney crisis, you can see a similar logic. Those who object to surge pricing, even those of us who understand it, would rather live in a world where people would do the moral thing – give someone a ride out of the danger zone, without making them pay through the nose. I know we don’t live in such a world, but I sure wish we did. And acknowledging that we don’t feels like admitting failure.

It’s one reason that I don’t pay my kids to do chores around the house. (The other reason is that I’m cheap.) I don’t want them to participate in the life of the family because of money but because it’s what family members do for one another and for the larger group. I’m not just trying to shield my kids from the realization that people are selfish; I’m trying to make them less selfish by eschewing market logic at home. Research has shown that markets don’t just force us to confront our selfishness, but often make it even worse.

Uber’s defenders have failed to acknowledge that this is a perfectly reasonable reaction. … you’d think Uber, whose business relies on being able to attract more users and more drivers, would try to figure out a different system that didn’t rub so many the wrong way.

OK, there is a literature based on why we think some transactions are more, or less, acceptable as market transactions. I don’t see the relevance here. In the Israeli day care case, the center found that people were all right with being late if it were treated as a market transaction (and so no guilt), but not if it were a non-market interaction. Sullivan feels the same way (and I share this feeling) about household chores. But Uber is anything but a non-market organization. Rides are market transactions. And I don’t see how it is reasonable for someone in want of a car to say to a driver ‘you owe me this’. Sullivan needs to fill in some spaces in this argument.

And at Demos we have Matt Bruenig. In response to Olivia Nuzzi’s claim about higher prices bringing out more drivers he writes:

This is the most problematic part of surge pricing that always goes unnoticed. In an equal society, jacking prices when demand outstrips supply causes those who need/want the rides the most to get them. In an unequal society, jacking prices when demand outstrips supply causes those who have the most money to get the ride. Despite the taunts of irrationality you read from journalists, it is actually totally “rational” for someone on the bottom end of an unequal economic distribution to prefer keeping the prices low, thereby allocating the rides by lottery, rather than allowing them to be jacked up, thereby allocating the rides by wealth.

His error here is to think that it is only when there is a price ‘surge’ that goods go to the highest bidders. But goods and services always go to the highest bidders. Admission to the Art Institute of Chicago goes to those who are willing to pay $18 or more to get in, and does not go to those who are not willing to pay $18. This is not a situation unique to Uber or to dynamic pricing – it’s everywhere.

A long post; where am I going with this?

When goods and services are priced such that supply meets demand, then those who are willing to pay the most for any particular good are the ones who end up buying it. That’s true for all goods and services that are sold for money, and it’s true whether a price is rising or falling. To criticize dynamic pricing because in a ‘surge’ it is is only those willing to pay the price who get the good or service, is to criticize the price system at its core. That’s not unheard of, but it is a very radical stance, and should be recognized as such.

And so I am going to back Uber here. Prices of scarce goods should rise in emergencies, as they already do in ordinary situations when excess demand arises, to encourage more supply, to encourage those who can find a substitute to use it, to get the goods to those who value them most. If I really needed a ride somewhere, I would rather pay an inflated price than find there are no cars to be had at all.