Resisting the Coalition of the Invisibles
*This article was first published on the VUB Chair in Surveillance Studies Blog, June 11th 2021 and is inspired by a seminar on workplace surveillance that took place on March 19th 2021. It implicitly quotes and paraphrases speaker Kirstie Ball*
Following a UK Supreme Court decision ruling that Uber drivers are in fact workers, Uber announced March 16th 2021 that its 70,000 drivers will at least receive the minimum wage, a pension and holiday pay. This kind of legal mobilization can be considered the ultimate stage of resistance to algorithmic control, rooted in conflicts of interest that arise when a growing sense of unfair treatment coupled with little room for remedy cause anxiety and precarity for those dependent on equitable treatment in the labour process.
Surveillance in platform work – which is controlled algorithmically – differs from traditional workplace surveillance on some crucial points. While both serve as an organizing process facilitating the employment contract, surveillance in platform work is characterized by a reductionist calculative rationality. Opportunities present themselves based on past performance and customer ratings. Instead of a conversation about personal development or training, poor performance and bad ratings result in less work allocated or – in the worst case – account deactivation. This situation leads to the hyper-meritocratic law of the vital few whereby 20% of employees get to pick most of the cherries to the detriment of the remaining 80%. Another crucial point of difference is that algorithmic decisions are usually not accompanied by an explanation or motivation. Opportunities to resolve problems quickly are rare as people behind the algorithm are hard to get hold of. In short, surveillance in platform work is totally pervasive and not easily challenged or negotiated.
As the processes behind the algorithms are opaque and the platform operators reluctant to take on responsibility as an employer, the coalition of the invisibles opens itself up for resistance. It has been shown that platform workers manage the way in which they are seen by the algorithm by developing an algorithmic imaginary in an effort to test what it is they are dealing with and to assess the terms of engagement. This can be done either by pacifying the algorithm and not getting noticed, or by studying the logic of the system to get a chance at the most lucrative work opportunities – think of Uber drivers hanging out at places where they will most likely get a good deal. The terms of engagement will be tolerated as long as these align with their own interests. In the opposite scenario, workers resist the algorithmic control and develop counter-strategies to re-align the terms with their own interests.
Research shows that resistance to surveillance in platform work follows the same pattern as resistance to surveillance in traditional working environments. Dubbed ‘algoactivism’ by Kellogg et al., different forms range from individual practical action (think of eking out space for agency by bending the rules; not cooperating with the rules) to platform organizing (online fora such as Turkopticon, Fermondo, Dynamo where platform workers find each other to share knowledge; or The Rideshare Guy, a blog about how drivers can maximise their income in car sharing marketplaces) to discursive framing about fairness, accountability and transparency (evolving from raising awareness about algorithmic bias to calling for professional codes of ethics) and finally to legal mobilization – The UK Supreme Court Uber decision is the result of a strategy on the ‘collective action’ side of the resistance continuum.
 Kellogg et al. (2019) describe the way algorithms control work in terms of 6 R’s: restricting behaviour, recommending courses of action, recording activity, rating activity in real time, readily replacing people, rewarding and gamifying high performance.
 For further reading, see Bucher, EL (2020) Pacifying the algorithm – Anticipatory compliance in the face of algorithmic management in the gig economy. Organization DOI: 10.1177/1350508420961531
 For further reading, see Shapiro, A (2018) Between autonomy and control: Strategies of arbitrage in the “on-demand” economy New Media and Society 20(8) pp 2954 – 2971
 Kellogg, K, Valentine, M and Christin, A (2020) Algorithms at work: The new contested terrain of control Academy of Management Annals 14 (1) pp 366 - 410