Tim O’Reilly just wrote a great post suggesting that in labour marketplaces, like the one Uber operates, algorithms might offer a better deal to workers than regulation. It’s a complex argument and Tim’s long post is well worth a read if you have the time. For those of you in the mood for a summary the centrepiece of his argument is that Uber allows market forces to determine prices and allows individual drivers to determine when and where they work. Those two features combined result in a more efficient system that works out much better for drivers (and everyone else) than it would if regulators force them to make their drivers employees. As Tim notes, top down micro-management of low cost work-forces to match supply and demand have been ‘a disaster for workers’. Think Walmart or McDonalds.
There are a couple of other factors at play here too. Firstly Uber has expanded the market – people take more cabs because it’s cars are reliably available at short notice. Secondly Uber drivers are in charge of their own lives in a way that salaried employees never are. Uber drivers decide when to work, and where to work. Moreover, Uber’s surge pricing helps them with that decision. As Dan Pink highlighted autonomy is a major driver of job satisfaction.
Nothing is wholly good or wholly bad and there is, of course, a flip side to this. Uber’s algorithm sometimes causes problems, including one case of gouging passengers during a crisis, and overly aggressive behaviour from Uber execs certainly makes one pause for thought, but these seem more like problems that can be ironed out than flaws in the model.
Stepping up a level, ‘Algorithms beat regulation’ is a very radical idea and Tim’s article is a call for academics and policy makers to start thinking about it seriously.
Superficially it sounds to me like it could work and therefore merits investigation. One of the most obvious challenges is that labour marketplaces have huge economies of scale and could end up as natural monopolies, requiring a whole new set of interventions from the regulator. Another obvious challenge is to understand whether the Uber model works in the same way outside of taxis.
In the short term, if it is true that labour marketplaces can offer a better deal for both workers and customers then we should expect the model to become much more prevalent. It’s interesting that despite massive investment in the sector not many of the other labour marketplaces has taken off like Uber has, and there are rumours that a couple of the larger ones are struggling to make their unit economics work. A thorough strategy piece analysing the different players and their models would be a very interesting read.
If this topic interests you check out Azeem Azhar’s email newsletter The Exponential View. It’s where I saw the link to the O’Reilly article.