Uber's tough market design challenge

1-CfBhhh2cCTl3vnbE4diu3AUber has a difficult challenge. Faced with voracious interest group attacks, every misstep is a potential public relations problem. Previously, they have been hit hard regarding ‘surge pricing’ whereby in periods of high demand prices may be several multiples their usual level. We economists like surge pricing because it effectively rations demand while encouraging supply. In order words, it puts the ‘market’ back in taxi market which traditionally had no price variation. At the same time, consumers — as they have always been — are worried that large price increases represent gouging and create uncertainty. (Of course, again economists suggest that some forward price contracts could assist here but Uber has yet to rise to that challenge.)

This week saw the other-side of the equation hit in: what if the downside for drivers might be too much because of an increase in supply? From The Verge:

But this Valentines day, while traveling through San Diego in an Uber car, Lane heard something that disturbed him. “The driver had a Ford Sync system, and it read his text messages out loud.” The message, which came wedged between numerous texts about a promotion for free roses, said, “UberX is very close to SURGE. It’s Valentine’s Day! People will be out all night and we didn’t activate new drivers to make earnings even higher this weekend.”

Now what this sounds like is that Uber deliberately withheld potential supply that might otherwise have been available so as to send a signal to drivers that a surge would occur. Uber claims miscommunication (something that it has persistently had issues with):

“The message was a poor choice of words by the local team but its clear purpose was to get more supply on the system, not less — to keep surge pricing down and the numbers speak to that,” says Noyes. “Only 3.1% of all Valentine’s Day trips in San Diego had surge pricing. The average Valentine’s Day trip price was 2% more expensive or $0.26 more expensive on average.  In addition there were 306 drivers onboarded in the 2 weeks leading up to Valentine’s Day.”

Noyes explained the bit about not activating new drivers to The Verge:

“[Noyes] explained the text simply noted that Uber did not onboard as many San Diego drivers as they could have that week because in the two weeks prior, a very large number of new drivers were added to the system,” The Verge’s Ben Popper writes. “Earnings had been low, and the company wanted to reward new drivers with a strong holiday paycheck.”

It seems sensible that you don’t want wild swings in supply and if that is under your control you need to manage it. But this highlights the double edged sword Uber is dealing with. It needs flexible pricing to ensure market balance but at the same time, stability requires some management and coordination. Had Uber opted for a hands-off approach, new drivers may have had a bad experience and left while existing drivers might feel slighted in curtailing their own Valentine’s plans when the predicted price increase didn’t come.

I love Uber and use it because there is always a car available in 5 or so minutes. This was actually a consideration in our family’s choice to move to a single car for the first time in two decades. The problem here is that, despite rhetoric, markets are not perfect in solving Uber’s fundamental challenge of ensuring there is always a slight surplus in the market — that is, supply has to exceed demand in any area in order for there always to be a car available in 5 minutes. That is, Uber wants a slight amount of inefficiency where otherwise efficiency would on average balance supply and demand. Uber’s model is extremely risk averse on the shortage side of the equation.

To achieve this, given that demand is variable and uncertain (to some degree), Uber needs to have a stable supply curve. This is because prices will have to adjust prior to their being a potential shortage. So that means that price rises have to be quick and they also have to be somewhat slow to come down. Moreover, the baseline price is a regulated price so long-term issues have to be regulated by quantity adjustments. So far from the market being a neat solution, it is actually a pretty challenging solution here because Uber’s preferences are far from neutral — it is targeting surplus not going for efficiency. Sometimes I wonder if Uber’s economic advisors really understand that.

This is not a new problem. Electricity markets operate the same way. Prior to there being markets, engineers used to forecast demand and adjust supply so that there was always a slight surplus; the idea being that you didn’t want consumers to have outages. Electricity markets changed that but they ran like a procurement auction — we need this amount of supply today and we will use bids to ensure we get it and also have a market price. Not surprisingly, while most of the time prices were pretty stable during periods of high demand, prices would spike. For those generators who were at the margin — that is, only came out to play when the prices were high — they needed those spikes in order to justify being built at all. (Similarly, for Uber surges will never diminish and always have to exist to ensure supply by entrepreneurial drivers). Now, standing between consumers (who had price uncertainty but love quantity certainty) and generators playing in this market were retailers. Retailers absorbed the price spikes (or bought insurance against them) and then charged consumers a stable price higher than cost to compensate for the uncertainty.

Uber faces the same challenge but has opted for (a) algorithms to estimate (guess) price fluctuations ahead of the market — perhaps an hour ahead; (b) prices set by regulation even if done dynamically (to ensure they don’t fall too fast) and (c) no one standing between the consumers and the suppliers (no insurance contracts). Thus, there is no product available for a consumer who wants quantity certainty and price certainty. Without that, anything that might suggest that price increases are not purely market driven (such as sensible supply management of new drivers), creates a problem and concern. Moreover, Uber has set itself up as a regulator but it is not independent.

What this suggests is that the best thing for Uber would be competition from other platforms. This is because Uber will be able to claim that if consumers don’t like what it is offering, they can choose another option. They actually have this by default from regulated taxis as well as independent limos. They should promote that more. Indeed, the true disruptive thing to do would be to interface with other platforms right on the app and show consumers their full options. This is the sort of thing Amazon did when listing other retailer prices on its own listings. Uber could do the same.

One Reply to “Uber's tough market design challenge”

  1. California regulators warned four app-based ride companies on Thursday that they haven’t provided some information required to receive state operating permits. The companies in turn said they are working to comply. The move came a day after several San Francisco supervisors blasted the regulators for not cracking down more on the transportation networking companies, or TNCs.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s