Platforms exist in two states - growth and steady-state. A mature steady-state platform maintains a stable and sustainable balance between supply and demand. But to create a platform, you have to build both supply and demand at the same time. Innovative platforms such as Uber are oriented towards expansion and growth - recruiting new passengers and new drivers, and launching in new cities.
|New Passengers||"Every week in London, 30,000 people download Uber to their phones and order a car for the first time. The technology company, which is worth $60bn, calls this moment “conversion”. It sets great store on the first time you use its service ... With Uber, the feeling should be of plenty, and of assurance: there will always be a driver when you need one." (Knight)|
|New Drivers||"They make it sound so simple: Sign up to drive with Uber and soon you’ll be earning an excellent supplementary income! That’s the central message in Uber’s ongoing multi-platform marketing campaign to recruit new drivers." (McDermott)|
|New Cities||"Uber has deployed its ride-hailing platform in 400 cities around the world since its launch in San Francisco on 31 May 2010, which means that it enters a new market every five days and eight hours. ... To take over a city, Uber flies in a small team, known as “launchers” and hires its first local employee, whose job it is to find drivers and recruit riders." (Knight)|
But here's the problem. In order to encourage passengers to rely on the service, Uber needs a surfeit of drivers. If passengers want instant availability of drivers (plenty, assurance, there will always be a driver when you need one), then Uber has to maintain a pool of under-utilized drivers. (Knowles)
Simple mathematics tells us that if Uber takes on far more drivers than it really needs, some of them won't earn very much. Furthermore, people with little experience of this kind of work may underestimate the true costs involved, and may have an unrealistic idea of the amounts they can earn: Uber has no obvious incentive to disillusion them. (This is an example of Asymmetric Information.) Even if the average earnings of Uber drivers are well above the minimum wage, as Uber claims, it is not the average that matters here but the distribution.
The myth is that these are drivers who can choose whether to provide a service or not, so they are free agents. Libertarians wax lyrical about the "gig economy" and the benefits to passengers. However, the UK courts have judged that Uber drivers work under a series of constraints, and are therefore to be classified as "workers" for the purposes of various regulations, including minimum wage and other benefits.
Uber has announced its intention to appeal the UK judgement. But if the judgement stands, what are the implications for Uber? Firstly, Uber's overall costs are likely to increase, and Uber will undoubtedly find a way either to pass these costs onto the passengers or to pass them back to the drivers in some other form. But more interestingly, Uber now has a financial incentive to balance supply and demand more fairly, and to avoid taking on too many drivers.
Uber sometimes argues it is merely a technology company, and is not in the transportation business. Dismissing this argument, the UK courts quoted a previous judgement from the North California District Court:
"Uber does not simply sell software; it sells rides. Uber is no more a 'technology company' than Yellow Cab is a 'technology company' because it uses CB radios to dispatch taxi cabs."However, Uber's undoubted technological know-how should enable it to develop (and monetize) appropriate technologies and algorithms to manage a two-sided platform in a more balanced way.
Why is the Uber ruling not quite as straightforward as it seems? @richardveryard @ricphillips pic.twitter.com/OXIgTHvA6z— Jeffrey Newman (@JeffreyNewman) October 29, 2016
Update: similar concerns have been raised about Amazon delivery drivers. I have previously praised Amazon on this blog for its pioneering understanding of platforms, so let's hope that both Amazon and Uber can create platforms that are fair to drivers as well as its customers.
Mr Y Aslam, Mr J Farrar and Others -V- Uber (Courts and Tribunals Judiciary, 28 October 2016)
Sarah Butler, Uber driver tells MPs: I work 90 hours but still need to claim benefits (Guardian, 6 February 2017)
Tom Espiner and Daniel Thomas, What does Uber employment ruling mean? (BBC News, 28 October 2016)
David S. Evans, The Antitrust Economics of Multi-Sided Platform Markets (Yale Journal on Regulation, Vol 20 Issue 2, 2003). Multisided Platforms, Dynamic Competition and the Assessment of Market Power for Internet-Based Firms (CPI Antitrust Chronicle, May 2016)
Sam Knight, How Uber Conquered London (Guardian, 27 April 2016)
Kitty Knowles, 10 of the biggest complaints about Uber – from Uber drivers (The Memo, 5 November 2015)
Barry Levine, Uber opens up its API – and creates a new platform (VentureBeat, 20 August 2014)
John McDermott, I've done the (real) math: No way an Uber driver makes minimum wage (We Are Mel, 17 May 2016)
Hilary Osborne, Uber loses right to classify UK drivers as self-employed (Guardian, 28 October 2016)
Aaron Smith, Gig Work, Online Selling and Home Sharing (Pew Research Center, 17 November 2016)
Ciro Spedaliere, How to start a multi-sided platform (30 June 2015)
Amazon drivers 'work illegal hours' (BBC News, 11 November 2016)
See further discussion with @wimrampen and others on Storify: Uber Mathematics - A Discussion
Uber Mathematics 2 (Dec 2016) Uber Mathematics 3 (Dec 2016)
Uber's Defeat Device and Denial of Service (March 2017)
Updated 6 February 2017