07 January, 2014

Elastic Collisions: The Limits of Dynamic Pricing

Early New Year's morning an Uber user paid $82 for a one mile ride. My first reaction to this news is that the rider appears doubly stupid: first for paying $82 for a one mile ride; and second for admitting to having paid $82 for a one mile ride. But the experience has kicked off a debate or at least a renewed awareness of dynamic pricing for goods and services.

Obviously the theory of dynamic pricing is to efficiently, some would say ruthlessly, match supply and demand for a given product or service at a specific time and location. In this regard, it's exactly analogous to how a *skilled* bookie uses the odds he sets to ensure he has as many bets for team A as for team B. So catching a cab at 1:47 AM on New Year's Day might be a high demand proposition. On the supply side of that transaction, I guess it's possible that some drivers might decide to be off duty at this time, thus lowering the supply, but I suspect the imbalance is mostly on the demand side of this particular transaction.

Balancing supply and demand assumes a certain amount of elasticity in both supply and demand. If the asking price of a service rises beyond a "threshold of pain" for a specific consumer, it is reasonable to assume that they will not engage in the transaction at that price. Personally, I'd have to be in an amazingly compelling scenario in order to knowingly and willingly pay $82 for a ride of one mile.

On the supply side, the elasticity is manifest in drivers' decisions to be out on the road offering rides. If a driver knows he or she can charge let's say 10% more than normal for a ride, some of them will find this premium attractive; others may decide that 10% is not enough to incentive roaming the streets at two AM on January 1st. Obviously, as the premium rises, so too does the number of drivers that will decide to enter the market.

The dynamic pricing model faces a limit to elasticity on both supply and demand sides. Above some price, the overwhelming majority of consumers will not buy. Below some price, suppliers will not find it worth their time to compete for customers. Obviously, this elasticity is in constant flux, is situation, and is temporal. When the IRIDIUM satellite phone system made its debut, I was asked who would be willing to pay more than $7 per minute to place a phone call. I assumed that two plausible answers were: a CEO who wasn't particularly sensitive to spending $7 per minute at any time and an arctic explorer who wished to call for help after a polar bear had bitten his leg off to whom money was no object at that specific time. My family goes through maybe two gallons of milk per week. Like George H. Bush, I will confess to being a bit hazy on the price I pay per gallon, but let's assume it's around three dollars. If it goes a bit higher, say to four dollars, our buying habits likely will go unchanged. But if the price suddenly went to eight dollars, I'd examine alternatives, including just shunning milk.

Efficiency in a market and how it sets prices (i.e., values things) is also limited by how promptly the links between supply and demand (and their respective elasticities) are mediated. The energy industry is an example of delayed market signals (or their realization) leading to unfortunate investment decisions. This is because while a cadre of Uber drivers can decide almost instantly to take to the streets, it can take a decade to carry out a plan to find and exploit energy resources.

The other limit to elasticity is information. Do prospective buyers know what the current going rated is and what the cost drivers are? Do suppliers know how many buyers are out there, how many competitors are in the game, and what strategies both buyers and sellers are using? An efficient and fair dynamically priced market also assumes a high level of information symmetry: buyers and sellers all have the same information upon which to base decisions. Automated systems like Uber ought to easily meet these criteria.

So why someone chose to spend $82 to be carried a mile escapes me, but I assume some other factor was at play. Maybe intoxication, seriously inclement weather, the desire to impress a date (or having done so, the urgency to whisk that date to one's apartment before the bloom falls off the rose)? All I can say is that if I willing paid $82 for a one mile trip, the destination would be a hospital and I'd be apologizing to the driver for all the blood emanating from my polar bear severed leg.

No comments:

Post a Comment