Pacifying the algorithm – Anticipatory compliance in the face of algorithmic management in the gig economy

Abstract

Algorithmic management is used to govern digital work platforms such as Upwork or Fiverr. However, algorithmic decision-making is often non-transparent and rapidly evolving, forcing workers to constantly adapt their behavior. Extant research focuses on how workers experience algorithmic management, while often disregarding the agency that workers exert in dealing with algorithmic management. Following a sociomateriality perspective, we investigate the practices that workers develop to comply with (assumed) mechanisms of algorithmic management on digital work platforms. Based on a systematic content analysis of 12,294 scraped comments from an online community of digital freelancers, we show how workers adopt direct and indirect “anticipatory compliance practices”, such as undervaluing their own work, staying under the radar, curtailing their outreach to clients and keeping emotions in check, in order to ensure their continued participation on the platform, which takes on the role of a shadow employer. Our study contributes to research on algorithmic management by (1) showing how workers adopt practices aimed at “pacifying” the platform algorithm; (2) outlining how workers engage in extra work; (3) showing how workers co-construct the power of algorithms through their anticipatory compliance practices.

Exploring the dark and unexpected sides of digitalization: Toward a critical agenda

“…Bucher et al (2021) study digital work on the platform Upwork. The authors detail how workers try to anticipate the decisions of deployed algorithms on the platform, and the practices they engage in to adapt to the assumed automated decision-making, to remain successfully employed.…”

“…At the present juncture, platform workers tend to be independent sub-contractors of platforms, which means that they miss out on any potential benefits of stable employment. As self-employed ‘entrepreneurs’, they are subjects to opaque algorithmic control, which leaves many of them, as shown by the study by Bucher and her colleagues (2021), with little insight into how their work is evaluated, what competition they face, and why they receive a work assignment or not. This ultimately leaves them to resort to additional efforts to stay on top of the algorithm, and equally minimizes their ability (and will) to collectively resist these practices, as stressed by Walker et al (2021).…”

Controlled by the algorithm, coached by the crowd – how HRM activities take shape on digital work platforms in the gig economy

“…Digital work platforms, such as Upwork or Fiverr, provide an increasingly common workplace for millions of workers -programmers, designers, writers and many more -worldwide (Kässi & Lehdonvirta, 2018). These new forms of digitally mediated work (Barley et al, 2017) present workers at the same time with a high degree of flexibility and autonomy, while simultaneously shifting the power balance away from the workers due to new forms of control and surveillance (Bucher et al, 2021;Duggan et al, 2020;Kellogg et al, 2020;Wood et al, 2019). Accordingly, digital work platforms challenge classic notions of the relations between workers and the organization in two major ways (Friedman, 2014;Gandini, 2019).…”

“…Instead, workers are considered entrepreneurs, freelance contractors or independent professionals (Upwork, 2019;Fiverr, 2020), and platforms position themselves as 'neutral' marketplaces that mediate transactions between clients and workers (Duggan et al, 2020;Kuhn & Maleki, 2017;Meijerink & Keegan, 2019). Despite platforms distancing themselves from a traditional employment relation, they still seek to impose measures of control on workers to ensure proper work assignment and performance management (Bucher et al, 2021;Duggan et al, 2020). As a result, workers often compare themselves to employees (Petriglieri et al, 2019), and several court rulings were given in favor of workers gaining employment status (Duggan et al, 2020).…”

“…Organizational socialization is seen as a key element in HRM, as HRM has to "send signals to employees that allow them to understand the desired and appropriate responses and form a collective sense of what is expected" (Bowen & Ostroff, 2004: 204). Yet, digital work platforms, such as Upwork, often intentionally obscure the responses they seek from workers and leave it to workers themselves to form a collective sense of what is expected of them (Bucher et al, 2021;Burrell, 2016;Fieseler et al, 2019). The lack of socialization is perhaps best exemplified by workers being left in doubt about what is it expected of them, as one worker noted: "There are three great mysteries in the world: 1] Stonehenge 2] Pyramids 3] JSS".…”