Andrew Stewart, ‘Uber and Out? Regulating Work in the Gig Economy’

We hosted Prof Andrew Stewart to talk about crowd working in the gig economy. The new forms of work are not entirely new, nor are they entirely negative in their implications for workers. But they do shake up social structures and relationships of work, and this has important ramifications. In these systems, workers bear a greater share of the risks of work. Stewart worries about how many service providers, all across the economy, will be protected across a number of different interests: labour relations and collective bargaining, superannuation, unemployment benefits, sick leave, and so on.

Stewart refers to De Stefano’s 2015 report on crowdwork to provide a typology of work in the gig economy. De Stefano points out that in some crowdwork platforms, the intermediary is much more heavily involved in setting and enforcing standards, prices, and controlling other aspects of how work is performed.

In Australia, estimates suggest that under 0.5% of Australians work on p2p platforms more than once per month, although that figure is expected to grow: Minifie & Wiltshire, 2016. Stewart makes the point that we have not yet seen the surge in contracting or freelancing work that has been apparent in the US, for example, where 15% of the workforce at contractors. In Australia, that figure, which varies between 9-11% of the workforce, has not varied much in the last decade.

Stewart talks about Uber as a relatively limited example for analysis of the gig economy. It’s obviously a flashpoint for concern, and Uber has been subject to legal challenges around the world on the basis that its driver-partners should actually be classed as employees. But Stewart notes that it’s quite unique in the context of labour. Unlike in other gig economy services that displace employment relations, Uber does not. It is well established in Australian law that taxi drivers are ‘bailees’, not ’employees’ – and this has been confirmed by the Full Federal Court and the Full Bench of the tribunal. This legal principle, which means that taxi drivers are not entitled to the benefits of employees, is problematic in Stewart’s view, but very difficult to change.

Uber’s contract with drivers is predictable in the way it seeks to avoid setting up its drivers as employees. The way it does this, in Stewart’s words, is a ‘reverse contracting’ model. Rather than contracting drivers to provide services to Uber, the contract sets out the relationship that the autonomous driver hires Uber to provide the networking and payment services for the driver’s business. So the contract is set up to maximise autonomy of drivers, but Uber also reserves a lot of power: it has the power to determine whether payment should be made against any given ride, the power to change the contract at any time, the power to deactivate drivers, and so on.

The law

Are Uber drivers employees? Stewart notes that some judges are very strong on the need to look to the reality of the enterprise. On this basis, it is clear that Uber drivers are working for Uber. But other courts take more conventional views, and require detailed analysis of the multiple indicia tests – on which basis Uber would possibly be able to avoid a finding that its drivers were employees with careful contract drafting. We are likely to see a test case in Australia in the near future, but it is not clear how it will turn out.

Stewart reviewed the applicability of other legal frameworks to taskworkers:

  • Uber certainly owes duties under workplace health and safety laws, which apply to all ‘workers’, not just ’employees’. It is not quite clear that health and safety laws will apply to other taskworkers, but they are much more likely to apply than employment laws.

  • For discrimination laws, most anti-discrimination laws are unlikely to apply for technical reasons. So where ratings systems are, for example, biased on sexist or racist grounds, our laws are not well-tailored to apply.

  • There is very little chance that workers have any right to superannuation contributions unless they are employees.

  • Workers compensation regimes are not likely to apply, although they are broader in some Australian states than others.

The future

Stewart suggests it could be useful to think about the spectrum of work in the context of the distinction between brokering services and labour hire agencies. Brokering services don’t exercise ongoing control at all, and may therefore might not attract liability or further responsibilities. Conversely, where a service like Uber that exercises heavy ongoing control about how work is performed, they should likely bear greater responsibility.

The more radical option is to completely reimagine our existing systems – away from relationships of employment, redesigning all these systems to focus on work. Or we could completely rethink the link between social benefits and work status (e.g. universal basic income).

The only option that doesn’t make sense, Stewart suggests, is to do nothing. We don’t want to allow employment to be displaced by ‘crowd-based capitalism’ without any adjustment of labour regulation and social safety nets currently tied to employment status.