The EU Directive: Worth Further Reflection

by Maria Antonietta Grasso, Naver Labs Europe — France

The research at Naver Labs Europe (O’Neill et al., 2013; Martin & al., 2014; Tuco & al., 2021) has shown that even in low skill platform work, the activity is embedded in a network of social exchanges. These social ties are required to assure work quality and to sustain the worker professionally and emotionally. Work is embedded in a network of relationships among members of the same community of practice and with a variety of business entities. This is true for every type of work and remains true also in gig work.

What is novel and generally appreciated by workers and all business entities using the platforms, is the flexibility it provides them in organising the work, while providing some stability in terms of demand and offer. However current platforms use a simplified model of work where the worker is made invisible and more similar to an element of a machine, rather than a human professional entity. This results in a number of issues that include lack of benefits, no professional progression, invisible workload, pay that can be lower than minimum wage, and being ruled by algorithms that can be opaque and over demanding. This is a new type of work organisation that is neither stable employment nor working on one’s own (Gray and Suri, 2019). In this context, the interventions of political entities, like the European Union providing a directive to regulate the sector of digital labour platforms, are required and welcome.

The directive intervenes in three areas: (1) the welfare protection (2) transparency of algorithmic decisions, and (3) the communication and collaboration support of the work environment.

In regards to welfare protection, research has shown that only a minority of platform workers rely on platforms as the main source of income. Gray and Suri (2019) report that roughly a power law rules this type of work, i.e. only a small percentage of workers belong to the always on stable category. The remaining ones are either exploring the possibility or already fitting such work into desirable lifestyle arrangements, taking into account its space and time flexibility. Given this picture, it is not surprising that the directive was fiercely criticised, e.g. in Italy, by some on-demand workers who were in fear of losing the possibility of the flexible arrangements they were enjoying. The directive should acknowledge these fears and aim at enactment of the welfare rights for the whole category of on-demand workers. As it currently stands, it seems that it forces this new type of work in the old stable work classification and it is missing to identify platform work as a new type of work that would require its own examination and regulation.

The second main chapter of the directive addresses the understandability of the algorithmically based decisions, an area where the workers today have to reverse engineer and accept management decisions without the possibility of a true exchange and transparency. The focus of the directive is on requesting the platforms to provide explanations clearly and in due time when requested. The directive demands to introduce “collective rights regarding information and consultation on substantial changes related to use of automated monitoring and decision-making systems”. This is certainly a change in the positive direction; however, it may turn out to be more a surface measure than a truly engaging one if we look at previous regulations about data management. A parallelism comes to mind with the GDPR right to request explanations about personal data use. Recent studies (Bowyer, 2022) show how such requests are answered with delay or not at all; and, more importantly, when provided they remain obscure and hard to understand for non-legal experts. A more impactful request would put in place human management channels granting human accountability about the management decision (the so-called human in the loop), instead of hiding the decisions beyond algorithmic formulas.

The third element is the request to provide a collaborative space for the workers. This is certainly an important first step, as it has been shown that workers communicate and collaborate to make the system work. They do this by physically joining in the same physical places to, for example, wait for food orders, and they connect to online fora to discuss and sustain each other. This happens for example to onboard newcomers, or to provide each other important advice and knowledge about how to carry out the work. The directive request is a first step, which should be followed by an accountable recognition for the time and effort this hidden work takes. Additionally, communication is essential to smooth performance of work, not only among workers, but with platform management, and service stakeholders. Thus, communication channels require explicit support and action, which is not considered in the current directive.

Finally, it is important to highlight that not only platform work will become more prominent, but that the algorithmic management enabling it will be more present in stable jobs, impacting also permanent employees. It is required and important to understand what rights are central to the ethics of work across the board.

Overall, the directive is important in taking a lead on regulatory steps and taking platform work more in the foreground. What is further required, however, is a larger step in producing regulations that fully account for the novelty of these work relationships. The current directive seems to push the new forms of work back in the old framing in order to protect them. A deeper and fresher recognition of their new reality may be required.