Your Private Driver: Surging the System

This is a weekly column that offers news, insights, analysis, and user tips for transportation network company (TNC) platforms like Uber and Lyft. Well, usually weekly, but the author has been somewhat preoccupied with a new job, new intensive schedule, and an upcoming move. He apologizes for the lack of content updates.

A new study has been making headlines this month claiming the Uber drivers have been finding ways to game the system to force riders to pay inflated Surge fares. Completed jointly by Warwick Business School and New York University, it claims to have found evidence that drivers are using tactics like logging off en-masse to artificially reduce the supply of available vehicles relative to demand, triggering Uber’s Surge algorithm to go into effect.

Fortunately, avoiding this problem is pretty easy: do nothing. Seriously, the results of this study are complete bull crap, and it’s filled with misinformation and other data that appears to be just completely made up. Sure, drivers have long talked about gaming the Surge system in this way, but none of their efforts have ever gained enough traction to become more than just talk. UberPeople.net, one of the largest and most well-known driver communities, is probably the most popular place for discussions of this type. Their userbase, however, barely accounts for a fraction of the number of registered Uber, Lyft, and other TNC drivers. Uber alone has over 1.5 million active drivers globally, with about 200,000 of them in the United States. I’ve yet to see any indication that a “log off and let it surge” collective action has ever actually happened, and even if it did, I’m not completely sure it would be successful.

In the conceptual stage, such strike attempts fail to take into account the vastly diverse needs and expectations of the many, many TNC drivers out there. A full-time driver, for example, would have little to lose from turning off their app for some time while waiting for more profitable conditions. A part-time driver, on the other hand, likely has only a few hours a day or week that they can drive, so sitting around with the app off would be a waste of valuable time for them. More than half of all Uber drivers (and I assume a similar percentage for Lyft) drive less than 15 hours per week. They want to spend that time giving rides, not wasting time on games.

It should be no surprise then that strike attempts also suffer from a lack of participation. At most, I’ve seen about a few dozen drivers willing to commit to any one single “app off” action. In Los Angeles, where there are literally thousands of TNC drivers available at any given time, a few dozen cars vanishing from the available pool isn’t likely to make a significant enough dent in supply or demand to trigger a Surge.

But what if they could? Let’s assume that a coordinated strike in a small area, like a popular nightlife district, was effective enough to dry up the supply and cause a Surge. The first thing I could see happening would be that other drivers unaffiliated with the strike would swoop in and snatch up waiting customers before the Surge rates got too high, making the strike a waste of time. Secondly, the riders could just wait the Surge out. Unless there’s a special event happening, Surges rarely last for more than a few minutes at a time. Patience wins the day.

Thirdly, and in my opinion the most likely scenario, would be that the drivers would wait until Surge multipliers had risen to a high-enough level (around 2x) then they’d all go online to snatch up some of those high-value fares. Unfortunately for them, the Surge pricing has killed some of the demand, so there aren’t as many rides to go around, and the sudden jump in supply from all the striking drivers going online would eliminate the Surge entirely. They’ve attempted to game the system but instead just wasted their time.

The Warwick study appears to have collected its data by interviewing some drivers and browsing the blogs and websites of others. Unfortunately they don’t seem to have any idea of how Uber actually works from the driver-partner perspective, they just read and listened to a bunch of people complaining about it. For example, this line was in reference to drivers trying to cheat their way to more money out of UberPOOL:

Professor Henfridsson added: “Drivers also either accept the first passenger on UberPOOL then log off, or just ignore requests, so they don’t have to make a detour to pick anybody else up. They then still pocket the 30 per cent commission for UberPOOL, rather than the usual 10 per cent.”

This shows a very basic lack of understanding of the driver-partner commission structure (drivers keep 75 percent of the fare, not 10 or 30), and creates the false assumption that skipping additional passengers during a Pool trip somehow makes drivers more money. If you’re gonna do a study about a product or service, shouldn’t you at least attempt to properly educate yourself about how that product or service works?

In my opinion, the real takeaway is not that Uber drivers are trying to find ways to cheat the system, but that the system is unintentionally designed in such a way that drivers have more of a short-term financial obligation to compete with their co-workers instead of cooperate with them. Any time a driver says that a fare is not worth the money, ten more are willing to swoop in to take it from him. A strike would be merely a golden opportunity for scabs to make a profit. There’s too much incentive to not take any kind of collective action for such organization to ever truly be successful on a large scale. The true monster of the gig economy is the realization that so few that are a part of it can afford to take a short-term loss in order to ally with their co-workers to achieve a long-term gain.

How Uber, Lyft, and the tech industry in general treat the growing engine of the ever-expanding gig economy will tell me a lot about Silicon Valley values; namely if they’re real, or just social media fluff.

Sekani Wright is an experienced TNC driver working in the Los Angeles metropolitan area. If you have any questions you would like answered for this column, you can contact him at djsekani at gmail dot com, or on twitter and reddit at the username djsekani. Have a safe trip!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.