Workers, consequently, carry most of the risk, while platforms benefit from a lack of employee idle time and payroll taxes.

Although the independent contractor classification constrains platforms’ ability to direct workers about ‘how’, ‘when’ and ‘where’ to perform the work, there are various subtle and covert ways through which they seek to influence how work is performed.

While the work’s hyper flexibility is often hailed as the main benefit for workers, our research reveals that in practice it is much more constrained. In effect, to an extent the flexibility discourse seeks to mask platforms’ control over workers.

In Australia and elsewhere, who controls the way the work is performed is at the core of legal tests that determine worker status. Crucial in regulatory decisions about the status of platform workers has been the question of ‘where’ and ‘when’ the work is performed, whereby higher flexibility can indicate less control.

Partly because platform workers are able to log on and off at their discretion, the Fair Work Commission and Ombudsman effectively endorsed Uber’s ride-hailing model. In other parts of the platform economy the classification debate, however, is far from settled.

Our research on food delivery, for instance, found that while workers in theory enjoy the same flexibility to select the time of work, it is much more constrained. For instance, by consumer demand.

Through their digital infrastructures including apps, how the work is performed is also more tightly controlled than platforms usually let on.

We identified three distinct forms of control that delivery platforms such as Uber Eats and Deliveroo exercise over workers.

First, algorithmic management ensures platforms are able to constantly monitor and capture data on worker performance. This is subsequently used to control and nudge worker behaviour, an issue to which Australian regulators and tribunals thus far have given limited consideration.

Second, by creating information gaps such as withholding important details of deliveries at the time of acceptance, platforms constrain worker choice. Although purportedly micro-entrepreneurs, workers are unable to make informed business decisions like whether a delivery is economically viable, making a mockery of the notion of enterprise.

Finally, workers receive only limited information on how the captured performance data, including consumer ratings, affects work allocation and continued platform access.

By obscuring the nature of their performance management systems as well as the other controls, platforms can elicit worker compliance including indirectly influencing ‘when’ and ‘where’ work is performed, illustrating that the narrative of hyper flexibility is somewhat of a mirage.

Dr Alex Veen (The University of Sydney), Dr Tom Barratt (ECU) & Dr Caleb Goods (The University of Western Australia). This article is based on an article recently published in Work, Employment and Society: ‘Platform-Capital’s ‘App-etite’ for Control: A Labour Process Analysis of Food-Delivery Work in Australia’