Surge Pricing is Surging: It’s an Unfair Game

An interesting story was at the center of music culture the last couple of weeks:

“Last week, The Cure announced their first North American tour in almost 10 years. As part of their deal with ticketmaster/livenation, The Cure demanded that ticketmaster refrain from using a dynamic pricing model. A dynamic pricing model is where if demand for a certain act is very high during the initial sale period, ticketmaster will then raise the ticket price unilaterally. That is, depending on when you check out, the price for the same seats may fluctuate from between, $80 to $300 dollars, as ticketmaster's systems identifies hot shows, acts, and seats.”

Last week, I wrote about Kanye West and inventory liquidation. 

This week The Cure and dynamic pricing.

As if someone is randomly choosing a celebrity and an operational concept to see if I can write about it…. 

Challenge accepted. 

Now, The Cure is not any band. While I’m not their biggest fan, their album Disintegration and the song Lullaby, were among my favorite as a teenager…but that was in the 80s. Little did I know that they still exist, and even less did I know that they have opinions on surge pricing and dynamic pricing: 

“The Cure were against this practice, so as part of their deal with ticketmaster, they demanded static ticket prices (tickets that are the same price no matter how high the demand is).”

Well, that didn’t go all that well. While Ticketmaster didn’t charge different prices. They charged different fees, prompting the following statement from Robert Smith (the band’s lead singer):

“‘WE [The Cure] DIDN’T AGREE TO THE ‘DYNAMIC PRICING’ / ‘PRICE SURGING’ / ‘PLATINUM TICKET’ THING… BECAUSE IT IS ITSELF A BIT OF A SCAM? A SEPARATE CONVERSATION,’ said Robert Smith in a tweet on Tuesday. His words came after a fan posted a tweet, claiming they’d purchased four tickets for $80, but after Ticketmaster’s fees, they total ended up being $172, which is more than double the cost of the ticket’s base price.”

Dynamic Pricing is Getting Popular

The idea of dynamic pricing (or surge pricing, as it’s usually called in ride-hailing) is becoming more prevalent and getting more attention from customers and the media. A couple of weeks ago, the Wall Street Journal published an article discussing the emergence of surge pricing in more day-to-day activities:

“Mr. Yenni, a 42-year-old advertising executive, tried to book online in advance at AMF Boulevard Lanes, where he wanted to reserve two lanes for two hours at 3 p.m. on the last Thursday of the year during winter break. The price quoted by the website, run by national operator Bowlero Corp., knocked him over: $418.90. ‘This strikes me as outrageous for a pedestrian family activity,’ he said. Mr. Yenni emailed Bowlero, and received a response saying due to the convenience of advance booking, ‘the pricing for a reservation is different than our walk-in rates or any current advertised specials.’”

But this is not only true for bowling, it’s also the case for movies…: 

“AMC Chief Executive Adam Aron said in a tweet that the different prices were similar to the way tickets for sporting events and concerts were sold, and that the system would be a way to hold down prices for the nonpremium seats.”

…and golf courses: 

“Golf courses also are teeing up the practice. ‘Dynamic pricing puts YOU in total control,’ says the website for the Saddleback Golf Club in Firestone, Colo., while noting the dynamic pricing rates ‘are constantly updating based on factors such as: the day of the week, the time of the day, Colorado weather, and most importantly golfer demand (or lack thereof) for specific tee times.’”

Dynamic pricing is as old as the notion of markets, when farmers would discount products toward the end of the day. But the notion of systematically charging every customer a potentially different price is associated with the airline industry. This practice has now become more prevalent as software tools allow firms to adjust prices continuously in response to demand, which is influenced by the time of day, as well as other factors. 

In all the examples above, the main message is that dynamic pricing is unfair. Why should different people pay a different price for the same product (even if we agree that it’s not always the same product)? In the airline setting, we got used to it. But not so much in other settings. This is, of course, related to the issue of priority queues, which I already discussed within the context of fast passes for ski lifts. 

But is Uniform Pricing Fair?

Uniform prices can also be unfair. The New York Times had an interesting article on the premiumization of everything: 

“Six Flags, the theme park operator, recently shifted to a more premium model by raising prices and limiting discounts, which Selim Bassoul, the chief executive, described as “bold changes to our business model in order to elevate the guest experience.” It has had mixed results so far. In the nine months through September, attendance at its parks fell by 25 percent from the year before, spending per guest rose 22 percent and, in the end, profits fell by nearly 10 percent.”

But again, we see a pushback:

“In January, the Walt Disney Company acknowledged that it might have pushed too hard on prices at its theme parks, angering loyal customers. It revised its policies on ticketing, hotel parking, ride photos and annual passes.”

The article warns against the gentrification of these activities, given that enough people are willing to pay the full price, choosing a uniform price (with fewer discounts) results in setting higher prices and thus pricing out a big part of the population. 

The Essence of Dynamic Pricing

Here’s what’s common in all these examples:

• scarce resources (bowling alley in peak hour, tickets for a concert, movies in desirable times, etc.)

• customers with a different willingness to pay for the service, and different time preferences. 

The combination of these two factors means that every solution will have flaws. Uniform pricing feels fair, yet it doesn’t allow people with time preferences to express them, and it also means that unless the price is very low, most people won’t have access to the service. 

Dynamic or priority pricing (which is what Six Flags does) allows people who have a higher willingness to pay for speed (or time) to express this preference while also potentially allowing the service provider to lower the prices in times that have less demand. 

The main issue, of course, is the negative externalities among customers given the limited availability. In other words, if enough people are willing to pay for a priority seat or the best time slot, those who can’t afford it are priced out. So maybe one relevant question is: When is dynamic pricing also fair? 

Now, this feels a little like beating a dead horse, so let’s add a little rigor to it. 

What is Fair Dynamic Pricing? 

Maxime Cohen (a co-author on several of my gig economy papers), Adam Elmachtoub, and Xiao Lei study the question of dynamic pricing fairness, particularly in settings where different parts of the population have a higher or lower valuation for a product or service than the general population. 

An interesting part of the paper is defining what fairness actually means. The paper studies four different types of fairness:

• Price fairness ensures that the prices offered to all groups are nearly equal. This is what The Cure were trying to do for their fans, and it’s what Six Flags was moving toward.

• Demand fairness ensures that access to the product is as close as possible across groups, meaning that the “prices should be set in a way that yields a similar market share for each group. For example, a local college may want to offer tuition loans or scholarships in such a way that each group has an equal probability of enrolling.”

• Surplus fairness requires that the surplus of the average person in each group (defined as the consumer valuation minus the price paid) is similar. This is not an intuitive notion of fairness, but it essentially means, “From each according to his ability, to each according to his needs.” 

• No-purchase valuation fairness implies that “the average valuation of consumers who do not purchase the product is approximately the same for each group.” In other words, the value lost in each group from individuals who could not afford the product should be similar. Again, not a very intuitive notion, but one that ensures that one group is not priced out of the market more than others. 

When reading the paper, it quickly becomes evident that fairness is not that simple. The notion of fairness is only relevant once we can agree on and identify the various groups among which we want to ensure fairness. The paper assumes that “customers are categorized based on an observable binary feature so that each group can be offered a different price.” In other words, to be fair, we must be able to identify the group association of each customer (e.g., race, income, gender, age, etc). If this isn’t possible, it’s hard to measure how fair the solution is. 

Can Dynamic Pricing be Fair?

The paper explores dynamic pricing and fairness by studying the impact of imposing the various types of fairness as constraints. Firms still try to maximize their profits via dynamic pricing,  but put constraints on how fair the solution is. The fairness constraint can be self-imposed or imposed by the regulator.

In the first (and in my opinion, the most interesting) result, the authors show that satisfying more than one of these fairness constraints is impossible. 

Impossible!

In other words, if everyone pays the same price, not everyone will have access, and the different welfare implications for each group will vary vastly. Some people are going to be priced out of the market, while others will get the product for a much lower price than they should. 

What if we impose only one of these fairness constraints? 

“we find that imposing a small amount of price fairness increases social welfare, whereas too much price fairness may result in a lower welfare relative to imposing no fairness.” 

In other words, requiring some level of price uniformity improves social welfare (which is probably what we should care about as a society). Interestingly, if you impose too much (for example, not allowing for any dynamic pricing), you get lower social welfare. Why? Because if the population size and valuations among groups are large, it’s better to target those who are willing to pay more, allowing the firm to maximize profits but pricing out a large part of the population, resulting in lower welfare overall. This is what we probably see with The Cure and Six Flags. 

What about the other types of fairness?

“On the other hand, imposing fairness in demand or consumer surplus always decreases social welfare. Finally, no-purchase valuation fairness always increases social welfare.”

So, if you want to make sure that everyone has access or that everyone has the same surplus (which means optimizing for “equity”), then the overall welfare will be worse. 

If you impose the “no-purchase valuation fairness” and ensure that the percentage of each class remaining out of the market is similar, then you always improve social welfare. But note that that requires different prices for each group. 

The implication: first, choose your fairness wisely. Choose the groups you want to be fair toward: gender, race, age, and income. And understand that there are trade-offs. When you choose a uniform price (price fairness), you are leaving big parts of the population out of the market. 

Every Resource Allocation is Unfair 

You may ask, should firms really maximize profits? That’s a whole different question. But it’s important to note that similar issues occur even when the service provider is not a profit maximizer and behaves as a social planner. These issues are very prevalent in health care. Vahideh Manshadi (my co-author), Rad Niazadeh, and Scott Rodilitz study this issue in a setting that was motivated by the Covid vaccine. 

But the issue of pricing and resource allocation is also common in educational settings. At Wharton, we give students the ability to bid on classes using a system developed based on a paper and the work of Gerard Cachon, called Course Match. Students allocate points to express their preferences. Course Match was introduced after the previous bidding system was viewed as unfair since people could “game it.” 

Second-year students have 5,000 points for the semester. First-year stufdents have 4,000 (as far as I understand). The outcome (from my own little tunnel vision): The 10:15 section of my “Scaling Operations” course cost 3,505 points, while the 1:45pm section cost 3,582 points. The 3rd and 2nd most “expensive” in Spring 2023 (I’m bragging, I know).

The question I hear from students at Wharton is that they pay tuition, which they believe gives them access to school resources, and now they don’t have access to a resource. Is that fair? 

As an instructor, I like the system overall since it means only truly committed students take the class. But it means that there are students that want to take it and can’t. I had a waiting list of a few hundreds of people (bragging again) Is this fair?

Oscar Wilde is quoted saying, “Life is never fair, and perhaps it is a good thing for most of us that it is not.”