Conference Agenda

The paper addresses social protection for platform workers at the European level and at the level of certain EU Member States.

The European Commission proposal for a Directive on improving working conditions of platform workers, which is currently going through the legislative procedure, seeks to improve social protection for persons performing platform work. The paper investigates the approaches to social protection in the draft Directive as well as in other documents issued during the legislative procedure. The most important questions are the provision of a rebuttal presumption of an employment relationship with binding force also for national social security institutions, the transferability of social security entitlements, voluntarily payments of digital labour platforms for social protection of platform workers and the extension of access to social protection to platform workers. Special attention is paid to social insurance against accidents at work and occupational diseases since safety and health at work were included by the ILO among the fundamental principles and rights at work. Furthermore, regulations and current/planned reforms at the national level in certain EU Member States concerning these issues are analyzed.

Until now, platform workers in the EU Member States generally do not enjoy social security protections. However, in certain cases platform providers voluntarily provide social benefits to platform workers. The paper compares such practices with manifestations of corporate social responsibility concerning social welfare. Changes in the business model of platform work (in relation to social protection) as a result of issued court decisions on employee/worker status of platform workers in certain EU countries are also taken into account. In addition, empirical studies concerning the reasons of platform workers for undertaking this type of work and what form of employment (employee/self-employed person) they would prefer are also investigated.

The paper adopts doctrinal legal research methods and provides a systematic legal comparison which takes into account existing empirical evidence in order to give concrete and detailed examples. The paper should contribute to the literature and the findings on how to extend and improve social protection for platform workers and on how to adapt social security systems to new forms of work and whether it is possible to opt for new regulatory approaches at the European level. The paper analyses the possible positive and negative effects of different solutions in law and practice.

Digital labour platforms are an important source of employment, with over 163 million workers registered on online platforms (Kassi et al., 2021). However, it is increasingly recognized that platforms have built precarious working arrangements, characterised by low pay, poor conditions, irregular contracts, opaque management practices and lack of representation, into their business models (Woodcock and Graham, 2020; De Stefano and Aloisi, 2022). Platform work largely operates outside labour standards, mainly because platform workers are generally classified as self-employed. While individual litigation has had some success in exposing sham self-employment, platforms are adept at reconfiguring relationships to avoid labour standards. This raises two questions. Firstly, how can labour standards be reshaped to meet platform workers’ needs regardless of employment status? Secondly, how can compliance with such standards be achieved? These twin research questions have been addressed by the Fairwork project. Since 2019, Fairwork has developed and applied a set of principles to evaluate the working conditions of platform workers in more than 300 platforms in over 30 countries, from a wide range of sectors, including ride-hailing, delivery, domestic, care and cloudwork. We interviewed thousands of workers and dozens of platform managers and gathered data from policy-makers and unions at local, national and international level. The principles, developed on the basis of wide consultation and extrapolating from existing international and domestic labour standards, have been tested and refined in the light of this data, to be fit for the purpose of regulating platform work. Simultaneously, we have developed alternative means of compliance based on research showing that informal sanctions, including negative publicity, can have as much impact as formal legal sanctions (Braithwaite, 2002; Paternoster and Simpson, 1996). Our project therefore works in two dimensions. First, we appeal to the influence of reputation and consumer power by publicly ranking platforms according to their compliance with our five carefully formulated principles based on fair pay, fair conditions, fair contract, fair management and fair representation. Secondly, we shape appropriate legal standards and explore avenues to achieve their formal adoption (Fredman, Du Toit et al, 2020). This paper presents our findings. We conclude by articulating the importance of regulatory intervention at international level and provide recommendations as to how ILO standards should be extrapolated to provide the right to decent work for platform workers.

Is adopting Data Mining and Machine Learning (DM&ML) methods in identifying labour law violations likely to support labour inspectorates to ensure the enforcement of labour law and keep up with rapidly changing labour relations? Inspection authorities have taken advantage of technological developments in collecting more and better-quality data and a wealth of data with the potential for a better understanding of who violates labour law, how they do it, and why they do it is stored in servers of numerous labour law enforcement institutions. Yet their prospects is rarely exploited towards improving their inspection rationale and strategies.

This paper glosses over why such data is not utilized. It focuses on revealing the superior predictive power of the DM&ML approaches compared to the manually configured red-flag approaches in targeting businesses for inspections and in increasing labour inspection’s efficiency to uncover undeclared work. Unlike manually set red-flag methods, the DM&ML ones increase the prediction accuracy by, first, automatically considering variables, that traditional paradigms of who, why, and how one engages in undeclared work or other labour law violations will omit and, second, identifying changes in behavioural patterns significantly faster than experts or practitioners.

In this study we demonstrate the application of a sophisticated and interpretable machine learning method, the Associative Classification, in the process of planning actions to face undeclared work and other labour law violations of the Albanian State Labour Inspectorate and Social Services. Interpretable machine learning produces ‘white-box’ classifiers presenting their results in explainable terms to humans, improving the users’ domain knowledge and their acceptability and trust in the models’ outputs. In our research application, we use actual data of around 40K onsite inspection visits performed around the country from mid-2019 to mid-2021. We build data mining models used in two ways: first, as an effective prediction tool for classifying risky employers, hence contributing to scheduling targeted onsite visits to deal with specific labour law violations; second, as a knowledge provision tool explaining to users how the predictions are made and revealing the most dominating employers’ feature patterns associated with the various labour law violations, thus enhancing the ability of the inspectors identifying these violations. We present the models’ classification outputs, their prediction assessment metrics, and paradigms of extracted knowledge. We prove that the proposed methodology using DM&ML approaches is highly more effective in several ways than the current inspection visits’ selection methods using red-flag indicators employed by the authority.