Wherever digital technologies collect and process data for the coordination of work, some form of algorithmic management is ubiquitously present in ways that significantly affect the organisation of work and job quality. But the algorithmic management of work is not new. It is an extension of a historical pattern of bureaucratisation of economic activity and the planning of work.
Once, the digital transformation of work focused on the impact of automation on jobs and on the digital skills demand. In the aftermath of the Covid-19 pandemic, digital transformation has now broadened its scope with concerns about the data that digital technologies can generate, collect, store, process, transfer and communicate. As such, digital transformation does not only impact the levels of overall employment, and the varieties of livelihoods and skills needed today, and in the future, but what has become more prominent is how such work is embarked on, coordinated, supervised and appraised.
Algorithmic management relies on digital technologies and data, and entails organisational and institutional choices on the specific use of those technologies for work coordination purposes. Algorithmic management is therefore shaped by socio-institutional and organisational factors. Collectively, these contribute to its shape and the productivity outcomes that are harvested.
Initially, the implications of algorithmic management were examined only within the framework of digital labour platforms. But, new ways of working have led to the extension of algorithmic management practices into factories, offices, hotels, ports, government services, and even wholesale warehouses. How novel aspects of algorithmic management interact with pre-existing organisational structures and topographies requires further interrogation. After all, it is the organizational culture, or company DNA, that determines its behaviour in the market. And this ultimately affects performance.
Uber offers the most noticeable example of the algorithmic management of work. Very often, gig workers are managed entirely by Apps that use algorithms. Uber drivers are not exposed to human management which is characteristic of a typical hierarchical work environment. Rather, it is the App that performs critical management tasks. From the cloud, Uber drivers receive job assignments, pay rates, performance management appraisals, and even suspension or termination from the platform. Algorithmic management of work evokes many concerns in relation to blameworthiness and authority.
In other settings, algorithmic management assists in making automated or semi-automated hiring decisions using input data about possible candidates. The work is searching for a worker. Not the other way around. The algorithms use various performance indicators to predict the best possible fit using Open Data sources like curricula vitarum, posts on social media platforms, the internet footprints of candidates, or even data aggregated from specially developed games.
Unlike gig-working platforms, these algorithmic tools still require a human manager. The algorithm does not take on the management function in its entirety. Infor Talent Management, for example, uses twenty-four behavioural characteristics to generate a data-driven predictive model that can “identify the best candidates”.
Data is an increasingly tactical and valuable economic resource. The collection, refining, and transfer of data in regular workplaces take us beyond the business model of digital platforms. Digital environments for all kinds of interactions have emerged as spaces where vast amounts of data can be collected from internet-based devices as well as from workers, clients, and customers using trackers and other digital tools.
This data may then be scaled for algorithmic management practices that are inconceivable. The velocity with which the data can be used to offer services and products, as well as for training machine-learning algorithms and for automated decision-making is a turning point for the emergence of what some have described as – “algorithmic panopticon”, or algorithmic surveillance.
Digital transformation affects work and employment through three vectors of change: digitization, automation, and platformization. Each of these vectors is associated with specific combinations of digital technologies that alter the way work is performed and organised. Although algorithmic management is particularly associated with platformization, it is enmeshed by all of them.
Algorithmic management is a combination of existing technologies that brings with it the potential to disrupt existing economic practices. Linked data analytics, AI, geolocation, and wearables are at the heart of algorithmic management. It is a specific way of using these technologies to automate some of the functions previously carried out by human management. In this sense, algorithmic management is a socio-technical process that adopts existing technologies to the institutional and organisation ecologies in which they are adopted.
Gig workers and the automation of work have caused employment law experts to turn their attention to the rise of algorithmic management. It is a shift that is characterized by the pervasive reliance on monitoring technology and algorithms to measure, guide, and sanction workers. It is a shift that has the potential to alter the legal regulation of labour markets. Automated management opens the spectre of granular employer control against a backcloth of opaque decisions made by algorithms in the absence of traditional management structures. It is a shift that portends a scattering of legal responsibility for every decision using cloud-based predictive talent analytics. The increasingly pervasive use of digital technologies across all economic activities, Open Data, and Open Finance make algorithmic management potentially disruptive for the future of work.
