An unfinished task? Matching the Platform Work Directive with the EU and international "social acquis"
Abstract
Besides straining international, regional and national employment status classification models, digital labour platforms are pioneering new strategies and approaches in terms of algorithmic management, digital surveillance, remote work and cross-border outsourcing, which are increasingly being adopted in more conventional sectors of the economy. Developments in the platform economy are thus crucial in providing a stress test for the resilience of existing labour standards, as well as providing useful input in terms of the reforms needed to ensure their suitability, the collective interest representation and mobilization aspects comprehended by rapidly changing labour markets.
This paper seeks to explore the key emerging regulatory dimensions of platform work. It contextualizes the challenges associated with platform work as an expression of the consolidated features that, in the past decades, have been transforming the labour market: non-standardization and the deregulation of employment relationships. Following that, it considers the definition of the personal scope of application as a key challenge faced by essentially all attempts to regulate platform work. It does so primarily by exploring the functions and operations of a legal device known as “presumption of employment”, currently being considered by the proposed EU directive on platform work as a key tool to address the complex employment status classification questions that have surrounded the “gig economy” since its emergence. The paper then provides a conceptual cartography of the various EU regulatory instruments (both existing ones and those currently in the legislative pipeline) that will, jointly, define the legal mosaic of labour rights applicable to the heterogeneous phenomenon of platform work in the years to come.
The paper suggests that recent regulatory developments reflect a persistent attachment to the dichotomous model of subordination versus autonomy. Even once the EU directive on platform work has been adopted, work relations in this area will not be exhaustively regulated by its provisions and other existing directives and instruments would still provide (and, in some cases, fail to provide) answers to various legal questions (such as the concept of working time, privacy at work and the information and consultation of workers and their representatives) that are central to the rights, and livelihoods, of workers providing their labour through digital platforms. The paper elaborates on the interlinkages, overlaps, and tensions between the EU’s regulatory instruments and identifies strengths and weaknesses, and potential areas for further elaboration and even legislative reform. This paper concludes that in order to improve the working conditions of platform workers, regulators need to rethink the traditional rigidities associated with the subordination paradigm.
Introduction
Digital labour platforms are at the centre of the debate on the future of work1 owing to their crucial role in advancing the use of technologies in mediating and organizing labour.2 The automation of organizational functions and labour intermediation, expanding the pool of available workers beyond geographical or company boundaries, has radically transformed business models, jobs and how work is organized and negotiated, challenging the relevance and effectiveness of existing ways of ensuring good (or at least decent) working conditions and incomes.3 Besides straining international, regional and national employment status classification models, digital labour platforms are pioneering new strategies and approaches in terms of algorithmic management, digital surveillance, remote work and cross-border outsourcing, which are increasingly being adopted in more conventional sectors of the economy.
Developments in the platform economy are thus crucial in providing a stress test for the resilience of existing labour standards, as well as providing useful input in terms of the reforms needed to ensure their suitability and the collective interest representation and mobilization aspects comprehended by rapidly changing labour markets. However, so far, the regulation of the many important dimensions of digital labour platforms has been lagging behind the curve. Regulatory interventions have mainly emerged as a consequence of, and in connection with, a number of successful litigation strategies predominantly promoted by trade unions and worker organizations. This has often happened in a haphazard and piecemeal manner, with victories in court being stifled by contractual changes unilaterally introduced by the platforms (e.g. through the introduction of “substitution clauses” to defeat employment status findings).4 It is only more recently, and on the back of these jurisprudential developments in the courts, that some national legislators and, at a supranational level, the EU have started engaging with the issue.
At national level, regulation has often approached the phenomenon narrowly and in a fragmented and casuistic way, typically by focusing on specific forms of work provided through platforms (e.g. couriers, urban transport workers) rather than holistically.5 By contrast, the EU directive, the draft of which was proposed by the Commission in 2021 and is currently under interinstitutional negotiation, aspires to be a more comprehensive intervention in the field as it introduces a number of rights and protections in principle applying to all those working through a digital labour platform, regardless of sector and form of platform work.
However, this working paper argues that even this approach may not be up to the task of regulating the broader phenomenon of “platform work”. Not only does its coverage risk being too tight, but the proposed directive also tends to neglect important aspects, such as the regulation of working time or access to collective bargaining, leaving these aspects to other EU instruments that may not be suitably up to date or sufficiently integrated with the more ad hoc rules included in the proposal.
This paper seeks to explore the key emerging regulatory dimensions of platform work. It is organized as follows. The next section contextualizes the challenges associated with platform work as an expression of two of the consolidated features that, in the past decades, have been transforming the labour market: non-standardization and the deregulation of employment relationships. Following that, it considers the definition of the personal scope of application as a key challenge faced by essentially all attempts to regulate platform work. It does so primarily by exploring the functions and operations of a legal device known as “presumption of employment”, currently being considered by the proposed directive as a key tool to address, and in many ways obviate, the complex employment status classification questions that have surrounded the “gig economy” since its emergence. The paper then provides a conceptual cartography of the various EU regulatory instruments (both existing ones and those currently in the legislative pipeline) that will, jointly, define the legal mosaic of labour rights applicable to the heterogeneous phenomenon of platform work in the years to come.
The paper suggests that, even once the EU directive on platform work has been adopted, work relations in this area will not be exhaustively regulated by its provisions and that other existing directives and instruments would still provide (and in some cases, fail to provide) answers to various legal questions (such as the concept of working time, privacy at work and the information and consultation of workers and their representatives) that are central to the rights, and livelihoods, of workers providing their labour through digital platforms. The paper elaborates on the interlinkages, overlaps and tensions between these instruments and identifies strengths and weaknesses, and potential areas for further elaboration and even legislative reform.
Are we ready for the next “pandemic of precariousness”?
Platform work did not emerge merely as a byproduct of the large availability of connected devices, shifts in consumers’ appetites and deep-pocketed venture capital funds. To tease out the implications of a trend towards the platformization or taskification of work relationships, one cannot help but look at the pre-existing factors and conditions that enabled their advent. They can be identified in the ineffectiveness of some areas of supranational and national labour law frameworks and, more generally, in the progressive decline in the quality and security of labour markets.6 These trends are often the direct result of an institutional preference for the flexibilization and deregulation of employment protection systems which, for much of the last three decades, has led to a progressive re-commodification of labour.7
This admission prompts us to situate platform workers within the overcrowded groups of non-standard workers that are falling through the cracks of the EU and national social acquis owing to their longstanding institutional strictures and rigidities.8 In basically all jurisdictions, workers still have to pass a “subordination test” in order to enjoy full employment rights, reflecting a regulatory model that was designed for a pre-casualized and pre-digital world of work. However, most forms of discontinuous, fragmented and remote labour, which make up a growing proportion of today’s work arrangements, may escape the classic notion of subordinate employment. And this challenge has been exacerbated by the rise of the platform economy.
Moreover, we argue that the exclusion of large groups of workers from enjoying decent work standards goes beyond the narrow definition of “who is a worker”. Even if these precarious forms of work were brought under the normative umbrella of an employment relationship, the existing body of labour norms would fail adequately to rebalance and compensate for the unprecedented contractual, control and economic power of employers. In many cases, this is a reflection of the very institutional design of some crucial aspects of international, EU and domestic labour law instruments which, we argue, are not suited to emerging models in which workers are geographically dispersed, workplaces are fissured, economic processes are conducted through a digital infrastructure and work consists of a series of continuous yet inconstant tasks.9
Fragmented notion of worker
Digital labour platforms have opened up the labour market to new “agile” business models, with a reconfigured notion of the firm and a strong reliance on outsourcing practices, and replaced traditional forms of task allocation and performance control with more subtle means of digital surveillance and algorithmic management.10 As a result, there is a growing mismatch between these forms of work and legal standards and types which are traditionally rooted in notions of vertical, hierarchical and rigid subordination. This, in turn, leads to a progressive inapplicability of traditional employment standards.
From a legal perspective, the success of the platforms in “escaping” their responsibility as employers mainly relates to two factors. First, their business model is based on contractual distancing between workers and their principal (employer), whereby the platforms hire predominantly self-employed workers and stipulate contractual relationships with intermediaries and subcontractors.11 Second, platforms have been mastering the art of tweaking contractual terms and conditions or slightly adjusting the business model to disguise subordination and mitigate the impact of court decisions and administrative orders. These alarming developments vividly demonstrate that, even after passing through the national sui generis “gateways” to labour rights,12 the actual improvement of working conditions is far from being achieved by platform-organized workers.13
The starting point of the conceptual cartography proposed in this paper is therefore that, in order to improve the working conditions of platform workers, regulators need to rethink the traditional rigidities associated with the subordination paradigm.14 Even the introduction of a rather broad presumption of employment, as proposed in the EU draft directive on platform work, will not solve all the wide-ranging problems related to the status of “worker”. This, we argue, is because this regulatory development reflects a persistent attachment to the dichotomous model of subordination (or direction or control) versus autonomy.
Certainly, through a legal presumption, workers can obtain a procedural simplification to vindicate the proper classification of their relationship by enforcing the well-established principle of the primacy of facts, as mandated by ILO Recommendation 2006 (No. 198).15 However, a further and more consistent broadening of the coverage of employment protection may be crucial to avoid further segmentation in the labour market.16 For example, as will be discussed in the next section, the directive proposed by the European Commission envisages that workers who can’t prove the existence of a certain degree of control by the platform won’t be able to trigger the presumption of employment, although they will still being able to enjoy a limited pool of rights (essentially transparency and information rights related to algorithmic management). The unintended consequences of such a fragmentary approach are not unknown.17
In the past, the incongruent regulation of new forms of work contributed to the consolidation of low-cost alternatives at the expense of more stable and better-protected forms of employment, justified by the salience of lowering barriers to job creation and developing flexible arrangements. Moreover, it is important to note that the fragmentation of labour protection is emerging not only as the result of a deliberate political and regulatory choice, but also of concomitant developments that have recently redefined the personal scope of application of specific labour rights. Reference can be made to the 2022 Commission Guidelines on the application of competition law to collective bargaining and self-employed workers, in which the European Commission recognizes the right to collective bargaining for a wide range of the (solo) self-employed.18 While this initiative is undoubtedly a positive institutional recognition of the right to collective bargaining for the solo self-employed, it also carries the risk of creating further segmentation among platform workers. For example, some platform workers may have the opportunity to negotiate their working conditions collectively, yet they may struggle to access individual labour standards unless they can prove that the platform exercises the necessary level of control to trigger the presumption of employment as set out in the directive.
At the same time, new paradigms are also slowly emerging in judicial decisions. Indeed, strategic litigation has gradually broadened its range beyond the classical issue of the appropriate classification of the work relationship. Claims have been brought before the courts in relation to working time, discrimination, health and safety, privacy and data protection.19 Here too, while the expansive interpretation of these labour norms reflects a positive transformative trend, it contributes to the fragmentation of the labour protections applicable to atypical forms of work and platform workers. Without coherent regulatory reform, judicial solutions have limited reach.
In the EU, regulatory action to harmonize and, at the same time, expand the scope of labour rights is required not only by the pressuring precariousness and increasing fragmentation of the labour market, but also by international labour standards. The ILO Committee of Experts on the Application of Conventions and Recommendations (CEACR) has repeatedly identified several national legislations that do not include casual workers, leaving them unprotected. Similarly, the ILO supervisory bodies have expressed concern that the exclusion of non-standard workers from the personal scope of national employment and labour laws presents obstacles to the application of the Fundamental Principles and Rights at Work (freedom of association and the effective recognition of the right to collective bargaining, the elimination of all forms of forced or compulsory labour, the effective abolition of child labour, the elimination of discrimination in employment and occupation, and a safe and healthy work environment).20
Effective exercise of labour rights
Guaranteeing formal access to labour protection to those in non-standard forms of work is a crucial first step. However, it does not per se entail that labour standards are effectively implemented and enjoyed. Again, the severity of regulatory (and interpretative) shortcomings is aggravated in the context of the platform economy. When the ILO Governing Body carried out a normative gap analysis on decent work in the platform economy, it identified a worrying number of aspects concerning working conditions in the platform economy that are not adequately regulated. These include non-discrimination practices, labour standards enforcement, employment security, decent remuneration, working time, data protection and, more generally, algorithmic management.21
This is hardly surprising since platform work barely fits into the normative concepts and paradigms that were conceived in times when work was done offline. Suffice it to consider that Directive 2003/88/EC on working time does not clarify how to account for time between tasks which is, therefore, mostly left unpaid. Another significant example is the exercise of the information and consultation rights enshrined in Directive 2002/14/EC, Directive 98/59/EC and Directive 2009/38/EC on European Works Councils. These instruments rely on the notions of “establishment” and “undertaking” which risk losing their meaning when labour is not provided in a physical environment and where business assets, such as data and algorithms, are predominantly intangible.
To these “historic” directives, EU legislators have added a second generation of labour norms, some of which are still in the pipeline, conceived when the wave of new challenges brought by the digital economy had already surfaced. These include Directive 2019/1152/EU on Transparent and Predictable Working Conditions (TPWCD), the GDPR, the AI Act and the draft directive on platform work itself. However, while attempting to regulate certain aspects of modern and flexible forms of work, we argue that these instruments also turn a blind eye to certain crucial specificities of platform work. They only partially address the existing shortcomings in the regulatory framework and do not provide tools which are sharp enough to improve working conditions in the platform economy. This reckoning urges us to focus on the more structural lacunae and shortfalls that prevent platform workers, as well as the legions of non-standard workers, from enjoying good (or at least decent) working conditions.
The drafters of the directive on platform work appear aware that only some of the vulnerabilities experienced by platform workers are addressed in the legislative text and have included several cross-references placing the directive within a complex mosaic of regulatory tools, implicitly admitting its insufficiency. Measuring the depth of these “protective gaps”22 therefore requires a thematic and combined reading of both old and new legislation. We believe that this exploratory exercise is essential in view of taking stock of the present and constructively looking ahead. In addition to providing a picture of the loopholes and solutions offered by the existing (and emerging) EU legal framework, this approach promises to boost the effectiveness of the draft directive, which cannot be read in isolation. At the same time, it aims to reduce the “corrosive” convenience of the platform paradigm, representing an open invitation to undermine working conditions in competing companies.23 Furthermore, we believe that the lessons learnt from the case of platform work should be used to define and implement robust regulatory and policy measures to prevent the next pandemic of precariousness, at a time when the virus of irresponsible and unsustainable business models could be infecting various economic sectors.
The legal presumption of employment: a tool but not a panacea
Platform employers have been making sustained efforts to obfuscate the employment status of their workforce through the use of, at times entirely fictitious, contractual devices (for instance through the use of “substitution clauses”) and through the deployment of technological innovations aimed at defeating established employment status tests (e.g. the possibility of switching between apps as an indicator of the existence of a broad “client base”, alongside the absence of “integration” into a particular business or of the assumption of “business risk”). Consequently, it is hardly surprising that a significant body of judicial decisions has rapidly emerged, in Europe but also internationally,24 around the question of the employment status of workers in the gig economy. It is also equally unsurprising that those advocating the fair regulation of platform work have identified the introduction of a “presumption of employment” as a possible and practical solution to this vexed question which may have the dual benefit of increasing legal certainty and reducing litigation (and litigation costs, especially for workers that are, more often than not, on low or very low incomes).
This section briefly explores the typical features of “employment presumptions” and then moves on to assess the presumption currently contained in the EU proposed directive on platform work as well as the, differently phrased, formulation proposed by the European Parliament in February 2023.25 It suggests that, depending on their specific formulation, “employment presumptions” can play a key role in clarifying the employment status of gig workers (and of workers at large), and can contribute to ensuring that platform work meets the minimal standards we associate with the concept of decent work. But it also warns against assuming that presumptions can be a panacea, or a silver bullet, singlehandedly resolving complex classificatory questions and tensions that, ultimately, are intimately linked to the concept and (legal definition) of employment that shapes the personal scope of the application of labour rights.
Legal presumptions of employment – typologies and functions
Presumptions of employment, broadly defined as statute-based or judge-made legal doctrines establishing that an unspecified work relationship is a legally relevant relationship of (usually subordinate) employment, are not an uncommon feature in a number of legal systems. Kullmann reports that, while “the majority of the EU Member States – that is, 17 – do not or no longer have a legal presumption regarding the employment relationship”, the remaining 10 EU Member States that do have a legal presumption either regulate general presumptions that are widely applicable to all kinds of working relationships (Croatia, Estonia, Greece, Malta, the Netherlands, Portugal, Slovenia, Spain) or inscribe presumptions that apply to particular groups of workers or sectors (Belgium, France).26
The ILO system contemplates the existence and usefulness of these devices, expressly mentioned in Recommendation 198 as possible tools for “the purpose of facilitating the determination of the existence of an employment relationship” and described as “providing for a legal presumption that an employment relationship exists where one or more relevant indicators is present”.
Depending on their exact formulation and design, legal presumptions can serve a number of different purposes and operate in different ways, with different degrees of efficacy in performing their primary function of facilitating the determination of the existence of an employment relationship. Abstractly speaking, there are different types of presumptions. They can be “absolute”; that is, non-rebuttable – very rare in the labour law domain but more common in other domains such as family and criminal law – and relative/simple presumptions, typically rebuttable by the other party. Some can be general (e.g. a general presumption that work is usually provided through a standard contract of employment27) while some can be specific (e.g. a presumption that certain professions (including journalism, modelling, work in the entertainment sector28) or forms of work (e.g. platform-mediated delivery couriers29) are typically performed by workers with a certain, protected, employment status. They can also be conditional in the sense of applying if and when certain indicators are present (for example, in the Belgian system), or generic; that is, applying to certain professions regardless of the presence of such indicators.
The primary effect of a rebuttable presumption is to establish that somebody (e.g. a working person, to use a neutral word) has a certain employment status until a different status is proven by the other party (e.g. the employer or principal). In that sense, presumptions have the incidental effect of shifting the burden of proof to the employer even though, conceptually and practically, “presumptions” and “burden of proof” are distinct legal institutions and devices. In most systems, this shift happens automatically as a consequence of the legal presumption (i.e. without the worker having to establish any particular facts) – a mere claim to being, say, an employee journalist on the basis of the presumption will normally suffice. But in systems where the presumption operates only after the worker has established, even just prima facie, the presence of certain criteria or indicators, for instance in the Belgian “presumption” referred to above, the presumption operates, in reality, as a mechanism for the reversal of the burden of proof conditional upon the claimant successfully establishing some basic facts that the lawmaker has identified as particularly indicative of the existence of an employment relationship.
By and large, general rebuttable presumptions (i.e. those applying to more than one form or type of work) and unconditional ones (i.e. those not subject to meeting certain criteria) are more effective than sector/job specific and conditional ones. The level of evidence expected of the presumed employer to rebut the presumption can also vary from system to system. And, as a general rule, systems where the employer can discharge the burden of proof associated with the rebuttal on the basis of more stringent thresholds (such as the “balance of probabilities”, as opposed to “sufficient prima facie” evidence – “beyond reasonable doubt” typically not being an option in civil litigation) will be more effective in protecting workers through a status presumption. No less importantly, and this is a point that will be developed further below, the effectiveness of legal presumptions will also largely depend on the breadth of the legal definitions of concepts such as “worker”, “employment relationship” or “employee” that are being presumed. By and large, the narrower the concept of “worker”, the easier it will be for a principal to rebut the presumption by claiming that the person does not, ultimately, meet the stringent tests and criteria used to define the “worker” (or “employee”) category.
So, while presumptions can indeed be very useful legal devices in the quest for clarifying the employment status of working people in general, and of platform workers in particular, the devil is often not only in the detail of how they are specifically designed and formulated but also in the ecosystem of the procedural rules and substantial definitions surrounding their functioning and operation. This is especially true of rules defining the burden of proof during the rebuttal and the concept of “worker” that, ultimately, the presumption will be tested against at the rebuttal stage.
The EU definition(s) of employment presumption in the gig economy
The EU has been a late adopter of legal presumptions in its labour law directives, to a large extent because, as noted above, presumptions can have significant, if incidental, procedural implications. Thus they sit uncomfortably with the EU law principle of “procedural autonomy” (a general principle of EU law whereby Member States are free to establish their own national procedural rules to govern the exercise of the law, subject only to ensuring that remedies for EU rights are equivalent to those applying to domestic ones and, ultimately, are effective in upholding them (the principles of “equivalence” and “effectiveness”)). However, even prior to the Commission proposal for a directive on platform work, the EU had deployed the use of presumptions in the employment context with the adoption of Directive 2019/1152 on Transparent and Predictable Working Conditions in the European Union, in particular its articles 11 and 15. This directive has significant implications for the rights of platform workers that are expressly mentioned in paragraph 8 of its preamble (provided they are not “genuinely self-employed”).
These two provisions contain two different types of presumption with distinct purposes and functions. Article 11 asks Member States that allow for the use of “on demand or similar” (e.g. zero hour) employment contracts to adopt measures aimed at preventing “abusive practices”; for instance, measures limiting the use and duration of such contracts or “a rebuttable presumption of the existence of an employment contract with a minimum amount of paid hours based on the average hours worked during a given period”. Article 15 introduces a presumption that is, in effect, a sanction on employers that fail to comply with the information requirements prescribed by other provisions of the instrument (that was adopted to substitute Directive 91/533 on an employer’s obligation to inform employees of the conditions applicable to the contract or employment relationship). It also asks Member States to ensure that, where a worker has not received in due time all or part of the documents referred to in Article 5(1) or Article 6, “the worker shall benefit from favourable presumptions defined by the Member State, which employers shall have the possibility to rebut”. Paragraph 39 of the directive’s preamble further clarifies that “it should be possible for such favourable presumptions to include a presumption that the worker has an open-ended employment relationship, that there is no probationary period or that the worker has a full-time position, where the relevant information is missing”.
While not being “presumptions of status”, both presumptions have the effect of introducing, by law, an employment contract or relationship where one, by contract, does not necessarily exist, including in circumstances where the employment status of the worker may be unclear. And, as noted above, these presumptions would apply to platform workers, as long as they are not genuinely self-employed, though inevitably their application would have to engage with a qualification of the employment status of the workers concerned (and we suggest that it would facilitate a finding of employment status whenever workers are given the opportunity to trigger it).
The Commission proposal for a directive on platform work
It is undoubtedly the legal presumption of employment relationship contained in Article 4 of the 2021 Commission proposal for a directive on platform work that has attracted most of the attention of commentators and experts.30 A “welcome instrument” but also one that under “the current formulation ... risk[s] being scarcely useful at best and even counterproductive in some cases”,31 the presumption contained in the original proposal has been substantially modified, even overhauled, by the European Parliament’s report which was adopted in February 2023.32 This section analyses both formulations, their merits and potential pitfalls, but with the caveat that inter-institutional negotiations (so-called 'trilogues')' are currently ongoing and that the final formulation of this crucial provision is, for the time being, far from clear.
In the directive originally proposed by the Commission, Article 4 was tasked with defining the conditions under which the “contractual relationship between a digital labour platform that controls […] the performance of work and a person performing platform work through that platform shall be legally presumed to be an employment relationship”. Article 4(2) set out five criteria and provided that the presumption would apply if at least two of them were satisfied. The criteria by and large revolved around the idea of the “control of the performance of work” and could be seen as descending from notions of “control” and “subordination”. The five criteria were: (a) effectively determining, or setting upper limits for, the level of remuneration; (b) requiring the person performing platform work to respect specific binding rules with regard to appearance, conduct towards the recipient of the service or performance of the work; (c) supervising the performance of work or verifying the quality of the results of the work, including by electronic means; (d) effectively restricting the freedom, including through sanctions, to organize one’s work, in particular the discretion to choose one’s working hours or periods of absence, to accept or to refuse tasks or to use subcontractors or substitutes; and (e) effectively restricting the possibility to build a client base or to perform work for any third party.
Much to the chagrin of labour advocates and trade unions, this presumption was not designed to operate “ipso jure” (i.e. automatically and by law) but could only “apply in all relevant administrative and legal proceedings”, effectively requiring a judicial or administrative assessment of the presence or otherwise of the criteria. Article 3 provided that, in ascertaining the existence of an employment relationship, decision-makers “shall be guided primarily by the facts relating to the actual performance of work”; in effect establishing a principle of the primacy of facts over contractual formalism, no doubt a positive and worker-protective approach.
In the original proposal, the Article 4 presumption was meant to work in unison with Article 5, the latter introducing the possibility for employers to rebut the presumption whenever “the digital labour platform argues that the contractual relationship in question is not an employment relationship as defined by the law, collective agreements or practice in force in the Member State in question, with consideration to the case-law of the Court of Justice”. Article 5 did not specify what burden of proof would need to be discharged for the rebuttal to succeed, but firmly placed the burden “on the digital labour platform”. Article 5 also offered the possibility to those performing platform work to argue that the contractual relationship with the platform is not an employment relationship.
The approach espoused by the Commission proposal has attracted some criticism. De Stefano pointed out that the five indicators contained in Article 4 were all linked to the concept of subordination and control and that, in most jurisdictions, each and every one of them (certainly criteria (b), (c) and (d)) could, on their own (i.e. without the need for additional criteria to be met), be used to classify an employment relationship as one of subordinate employment.33 It may also be argued that limiting the presumption to relationships that already display strong indicators of subordination and control over the performance of one’s work is not particularly helpful as those criteria would normally assist workers in claiming full employment status, at a full hearing, in most jurisdictions. This is partly a limitation arising from the fact that the proposal did not (and, on the basis of the principle of “procedural autonomy”, possibly could not) specify a lower burden of proof for the presumption to apply (e.g. a prima facie assessment) compared to the normal burden of proof deployed in employment cases (usually the “balance of probabilities”). It is also clear that these lists of indicators lend themselves to often unsophisticated, but none the less successful and effective, attempts by platforms to bypass them through the use of highly improbable (but not necessarily “sham”) clauses, such as loosely formulated “substitution clauses” to override criteria such as (d) or the tacit or explicit permission to use multiple platforms to override (e). In this sense, as pointed out by De Stefano, these articles of the Commission proposed directive are hardly “futureproof”.
The European Parliament proposals
The proposal of the European Parliament departs quite radically from the approach of the Commission’s text, most visibly by effectively deleting all the criteria contained in Article 4(2), with Article 4(1) simply setting out that, “The contractual relationship between a digital labour platform and a person performing platform work through that platform shall be legally presumed to be an employment relationship and therefore digital labour platforms shall be presumed to be employers”. Conversely, Article 5 is still offering the opportunity to digital labour platform employers to rebut the presumption “by means of demonstrating that the person performing platform work is genuinely self-employed” when they can satisfy the existence of both of the two criteria set out in the newly phrased Articles 5(3)(a) and 5(3)(b). The former provides that this could be when “the person performing platform work is free from control and direction of the digital labour platform in connection with the performance of the work, both under the contract for the performance of the work and in fact”; and the latter applies when “the person performing platform work is usually engaged in an independently established trade, profession or business of the same nature as that with which the work performed is related”.
The EP amendments to Article 5 went on to set out the indicators that courts and decision-makers shall take into consideration when establishing “control and direction” under Article 5(3)(a). These are the following: (a) effectively determining, or setting upper limits for, the level of remuneration or issuing periodic payments of remuneration; (b) effectively determining or controlling working conditions, including restricting the time schedule and working time duration, or enforcing the performance of work, including through penalties or incentives, restricting access to work or using rating systems as a tool of control and a basis for penalties and as a tool to allocate work assignments; (c) effectively preventing the person performing platform work from developing business contacts with potential clients, including via controlling or restricting communication between the person performing platform work and the recipient of goods or services during or after the performance of the work; (d) tracking or supervising the person performing platform work while carrying out the work; (e) requiring the person performing platform work to comply with specific rules with regard to appearance, conduct towards the recipient of the service or performance of the work; (f) effectively restricting the use of subcontractors or substitutes to perform the work; (g) effectively restricting the possibility of the person performing platform work to perform work for any third party, including competitors of the digital labour platforms; (h) restricting the freedom of the person performing platform work to choose social protection, accident insurance, pension scheme or other forms of insurance, including through adverse consequences.
While this list of indicators is reminiscent of the one deployed in Article 4 of the Commission proposal, this approach represents a significant departure from that originally espoused by the Commission. Firstly, by shifting the criteria to the “rebuttal” stage of the process, in effect it places the additional burden of meeting them on the platform employer, shifting it away from the worker claiming misclassification. Secondly, the indicators contained in Article 5 are suitably loose as would reasonably permit a wide range of persons working through platforms to claim that some degree of control and direction over their work is exercised by the platform (see the use of words such as “effectively” or the emphasis on “tracking” alongside the more obvious concept of “supervision”). Thirdly, these indicators do not apply cumulatively but separately, so a worker could ultimately argue that she or he was under the control or direction of the platform simply by setting out that one of them alone applies to the way her or his work is performed. Finally, platform employers are placed under a cumulative burden of demonstrating that both articles 5(3)(a) and (b) apply; that is, the worker in question is neither under their control and direction but is usually engaged in an independent trade, profession or business.
Sure enough, even this different approach lends itself to the criticism of providing a list of boxes that employers could attempt to tick simply by revising their contractual arrangements and introducing sham clauses or other defeating devices. But, overall, this exercise would prove much more demanding than the equivalent exercise they could perform under the Commission’s original proposal.
Legal presumptions and the “missing ingredient”: the concept of worker
The previous subsection offered an assessment of the legal presumption and rebuttal rule as formulated in the draft directive originally proposed by the Commission compared to the alternative drafting put forward by the European Parliament. It concluded that, on balance, the approach adopted by the European Parliament (an unqualified presumption, and a qualified rebuttal rule) is likely to be more worker-protective than the one espoused by the Commission (a qualified presumption with cumulative criteria and an unqualified rebuttal rule).
It is, however, worth noting that the worker-friendliness of these, no doubt important, procedural devices remains, ultimately, heavily dependent on the substantive definition of employment relationship that underpins their operation. Ultimately, short of introducing an absolute and non-rebuttable presumption of employment, any formulation of an employment presumption will need to be reviewed by an administrative or judicial authority. Their task, at some point, will be to verify and confirm that the presumed worker/employee is in fact and in law a worker/employee; that is to say that she or he meets the tests and substantive criteria used within that jurisdiction to determine whether a working person should fall within the personal scope of application of employment legislation.
Rules on presumptions can surely establish or facilitate the prima facie confirmation that a worker with a dubious or complex status is to be covered by labour rules. But ultimately this assessment will be reviewed and assessed against the standard definition of worker applying in that context. The narrower that definition of worker/employee, the more improbable it is that a decision-maker will confirm that someone working in a relationship that departs from the standard canons and criteria shaping that definition is indeed a worker or in an employment relationship. Conversely, the broader that concept, the more likely it is that a judge will eventually confirm that the presumption applies. Using an analogy from the field of zoological classification, one could say that one may very well introduce a presumption that dolphins are, really, chimpanzees. But as long as we retain a concept of chimpanzee with which we are all intuitively familiar, it is highly improbable, if not impossible, that any reasonable reviewer of that presumption will confirm that assessment. If, however, the chimpanzee ‘category’ were to be substituted with a broader concept (e.g. “mammal”, or “animal”), then the presumption would most likely receive the stamp of approval by most reasonable decision-makers.
And here lies, perhaps, the greatest shortcoming of all the presumptions discussed in this section. The personal scope of application of the draft directive on platform work (in both the Commission and the EP’s proposals) remains heavily shaped by a definition of “platform worker” that is defined as “any person performing platform work who has an employment contract or employment relationship as defined by the law, collective agreements or practice in force in the Member States with consideration to the case-law of the Court of Justice” (Article 2(1)(4) of the Commission proposal). Even assuming a slightly more generous approach in the Court of Justice (CJEU)’s formulation of the concept of “worker” compared to that in several Member States, ultimately the presumption will have to be tested between the rock of the national definitions of worker and the hard place of the CJEU’s jurisprudence. The latter, by and large, remains equally anchored to concepts of subordination and control, concepts that may not suit several types of work provided through platforms. There is a clear sense that, while some forms of platform work, for example those that are commonly associated with food delivery or urban transport, may more readily fit traditional, control/subordination based definitions of worker, other types, especially those performed solely online (as opposed to “on location” – noting that the draft directive purports to regulate both), are more likely to sit uncomfortably with these narrow and traditional concepts. This is the case even though those involved in these forms of work may well be providing work in a purely or predominantly personal capacity and earning all or most of their living through that work.
This is, perhaps, a shortcoming that has not received the attention it merits in spite of the draft directive, for the limited number of rights it introduces (essentially the ones contained in Chapter III), encompassing a broader scope of application than the one defined by the notion of “platform worker”. This broader scope stems from its attachment of those rights to the wider category of “person performing platform work”, defined by Article 2(1)(3) as “any individual performing platform work, irrespective of the contractual designation of the relationship between that individual and the digital labour platform by the parties involved” (emphasis added).
While the more recent jurisprudence of the Court of Justice suggests a slow move towards broader understandings of concepts such as work and occupation,34 it is fair to say that this remains the Achilles’ heel of this area of labour law and procedure. Ultimately a presumption will always be less effective if the concept of the employment relationship to which it refers is narrow and not sufficiently inclusive. A concept of an employment relationship that embraces an expanded idea of “personal work” would in many ways mitigate, if not eliminate, this type of shortcoming.35 It is worth noting that the 2022 Commission Guidelines on the application of competition law to collective agreements regarding the working conditions of the solo self-employed,36 which also apply to self-employed platform workers meeting the personal scope of the Guidelines, has toyed with the concept of “personal work”. The Guidelines are meant to apply to every “person who does not have an employment contract or who is not in an employment relationship, and who relies primarily on his or her own personal labour for the provision of the services concerned” (emphasis added). There would therefore be no shortage of examples of broader scopes to choose from should the EU decide, one day, to expand its definition of “worker”.
Legal avenues to improve temporary, poorly paid and unpredictable jobs in the gig economy
Bringing platform workers within the protective scope of labour law is undoubtedly a necessary and crucial step, yet is by itself insufficient to guarantee decent working conditions. Affording workers access to employment rights, for instance by means of the presumption of employment, grants them formal recognition to claim the application of national and EU labour law. However, the practical enjoyment of these rights can be seriously hindered by the challenges of translating labour norms designed in the pre-digital era to the platform economy and then implementing them.
A number of difficulties in applying existing labour law frameworks to platform work arise from the predominant business model in the platform economy which centres on the structural decomposition, fragmentation and reorganization of traditional production and economic processes. In other words, platforms are key enablers and accelerators of the disintegration of the firm as traditionally conceived.37 Concomitantly, the traditional spatial dimension of work has been deconstructed. In the conventional economy, notions such as the work environment, establishment and undertaking identify physical spaces where the employer has an obligation to ensure that workers are in a condition to exercise the rights that the legal framework confers on them (for example, information and consultation rights). In the context of the platform economy, instead, there is a predominance of virtual workplaces and a more dispersed workforce, which makes the application of labour norms quite challenging, especially for those rooted in the notion of the firm as a relatively static and tangible reality.
But the “spatial” dimension is definitely not the only thing being deconstructed. The “temporal” element is also disrupted since platform work aggravates the “work on demand” feature of current labour markets. It represents a throwback to the most precarious forms of casual work38 and is reminiscent of a production model based on labour intermediation and disenfranchised own-account providers of personal work.39 The extreme fragmentation of work into “disjointed episodes”40 (“gigs”), organized on an “if and when” basis and compensated in a pay-as-you-go manner, is instrumental in making work devoid of a wide panoply of employment protections and in the context of both business risks (for instance, the drop in demand) and costs (those for equipment, fuel, training, customer acquisition) being offloaded to such workers.41
Key sources of frustration for both on-location and online platform workers are the unpredictability of schedules, the short-lived nature of assignments, the lack of transparency in managerial patterns and the danger of one’s employment relationship effectively being terminated simply by an algorithm which is no longer offering “job” opportunities. This problem has a knock-on effect, leading to poor income levels and poor safety and health conditions. Caught in a vicious circle, workers cannot plan with a reliable degree of anticipation and are forced to compete with each other to secure a high number of well-paying “jobs” by appeasing the opaque algorithm-based systems deployed by companies. Such an agonistic version of work commitments exacerbates psychosocial risks and exposes workers to underemployment or overworking,42 sometimes owing to solutions such as the opening of accounts on multiple platforms (a factor that could be relied upon by platform companies, and possibly some courts, to weaken the application of the presumption of employment in its current formulation).
Finally, a significant challenge in effectively applying the existing body of EU and national labour law is the difficulty in assigning employer responsibilities and obligations to the appropriate level. Even more than on-location platform work, crowdworking epitomizes the digital fuel that labour platforms give to outsourcing and subcontracting practices.43 Within disaggregated firms, numerous contractual relationships for the provision of labour exist, yet there is a tendency in which no single entity assumes the full responsibilities of an employer. Borrowing this concept from Weil, the “organizational glue” that holds the economic process together lies not in traditional firm structures but in algorithmic management.44 Algorithmic management tools and practices in fact supervise, steer and integrate labour processes across the boundaries of conventional firms. Quite evidently, the EU legal framework (still) lacks adequate provisions to address and regulate such a disruptive change. This leaves the practices of excessive monitoring and pervasive management broadly unregulated and lacking clear accountability.
In this section, we present a selection of these protective gaps. We consider a number of policy areas (collective rights, working time, the transparency and predictability of working conditions and management by algorithms) with the aim of exploring the challenges of the existing frameworks and the possible responses they can offer.
Redefining the space for the exercise of collective labour rights
Collective labour rights play a crucial role in counterbalancing the employer’s stronger economic power and in promoting workplace democracy. These rights encompass collective bargaining as well as information, consultation and, in the most advanced systems, employee participation in relation to management choices and substantial changes in work organization. There is also a degree of functionality between information and consultation rights and collective bargaining, as the establishment of information and consultation channels and practices can create a fertile environment for the further negotiation of company level collective agreements.
Recent developments in EU policymaking suggest that platform workers should be entitled to the full enjoyment of these rights. As mentioned in the previous section, the 2022 Guidelines on collective bargaining for the solo self-employed clearly indicate that collective bargaining practices among genuine “solo self-employed persons working through digital labour platforms” are compatible with EU (competition) law. Moreover, the proposed directive on platform work includes information and consultation rights among the minimum standards to be recognized in respect of platform workers, as long as they are in an employment relationship. Therefore, in principle, platform workers are granted collective labour rights. However, can they effectively exercise them?
Information and consultation rights are one of the most regulated aspects of EU labour law.45 The first norms were adopted in the late 1970s, targeting the specific events of business restructuring and collective redundancies (now Directive 98/59/EC) and transfers of undertakings (now Directive 2001/23/EC).46 These directives aimed at establishing minimum standards to reduce the adverse effect of business reorganization on working conditions.47 A second regulatory wave followed in the 1990s and the 2000s, resulting in the adoption of a directive establishing a general framework for information and consultation rights (2002/14/EC) and one on European Works Councils (now 2009/38/EC).48 Unlike previous instruments, these norms did not focus on specific episodes in the life of the firm; rather, they considered information and consultation rights as integral components of the regular management of the company. Within this articulated framework, EU workers and their representatives have fairly extensive rights to be informed and consulted by management, provided that certain quantitative requirements n terms of the size of the workforce are met.49
Although the draft directive on platform work refers explicitly only to Directive 2002/14/EC,50 in principle all of the existing EU acquis on information and consultation rights would also be applicable to platform workers, at least to those in an employment relationship. In addition to the rights already enshrined in EU law, Article 9 of the Commission’s proposal requires the platform to inform and consult workers’ representatives (or, in the absence of these, workers themselves) “on decisions likely to lead to the introduction of or substantial changes in the use of automated monitoring and decision-making systems”. Furthermore, Article 15 demands that Member States ensure that “digital labour platforms create the possibility for persons performing platform work to contact and communicate with each other, and to be contacted by representatives of persons performing platform work, through the digital labour platforms’ digital infrastructure or similarly effective means”.
What the proposed directive does not consider or acknowledge, however, is that the exercise of these rights may be far less straightforward than it appears at first sight. Indeed, the proper implementation of information and consultation practices requires the identification of the “locus”, or unit, from which the workers’ representatives derive their representative legitimacy and obtain their mandate to negotiate, and this may be difficult to define in the platform work context. To identify this space, the existing EU directives mentioned above rely on the concepts of “establishment”51 and “undertaking”,52 to which dimensional requirements are then applied. Conceived in a pre-digital environment, these concepts imply identifiable spatial and organizational characteristics, while they also refer to a world in which the workplace has not yet been as dramatically affected by outsourcing trends, and it was fair to assume that all (or most of) those working in the same “space” have the same employer.53
The term “establishment” has been defined by the Court of Justice as the “unit to which the workers are assigned to carry out their duties”.54 The characteristics of this unit are that it is a distinct entity with a certain degree of permanence and stability and it has the technical means and a certain organizational structure which enables it to carry out the assigned tasks.55 At the same time, the Court has clarified that having a recruitment and dismissal department is not necessary56 and neither is legal, economic, financial or technological autonomy decisive.57
In the absence of strict administrative and operational requirements, managerial discretion in defining the boundaries of the firm’s establishments is quite wide, which is rather problematic from the point of view of ensuring effective collective labour rights. The risk of the employer organizing its business into a large number of units with the aim of keeping the size of the establishment below the dimensional requirements set out in the directives has already been encountered in conventional businesses.58 The danger is exponentially greater in the platform economy, where the company’s structure is wholly or in part located in a digital ecosystem, the business’s organization is fluid and there is no physical workplace.59
The space for the exercise of information and consultation practices and related collective rights therefore needs to be redefined.60 This may be relatively easy for on-location platforms (such as those providing food delivery, care, handyperson or courier services) where, even in the absence of directly identifiable premises, the execution of the work has a tangible spatial dimension.61 It is no coincidence that a recent empirical study on the working conditions of food delivery riders in Madrid showed that the workers consider the street to be their actual workplace.62 The notion of “establishment” could consequently be reinterpreted according to geographical and topographical criteria, in terms of cities or specific districts.63 This would not be entirely new in European legal systems. As Donini has pointed out, the Italian Court of Cassation has ruled that the place of reference for the exercise of the collective rights of door-to-door sales representatives is the territorial area in which they carry out their activities.64 Moreover, these territorial forms of representation seem well suited to giving a voice to workers who are active on more than one platform at the same time.65
For online gig work, however, the situation is more complex. To begin with, it may be more difficult for these workers to activate the presumption of an employment relationship established in Article 4 of the directive as proposed by the Commission. As discussed above, the reversal of the burden of proof is based on a set of criteria anchored in a traditional notion of control that does not fit well with the more subtle (but not as a result less ubiquitous) type of monitoring carried out by microtask platforms. This may mean that they will have to resort to more indirect means to assert their collective labour rights, for instance by appealing to the universal scope that ILO Convention 149 (No. 98) confers to the right to collective bargaining.66 They may also benefit from the recently adopted 2022 Guidelines, with the caveat that this is a non-binding instrument and its effectiveness may be limited.
Moreover, unlike in the case of on-location platform work, the element of territoriality in the execution of work completely loses its relevance when the work is performed fully online. One way to overcome the lack of a tangible environment could be to link the exercise of collective rights to (digital) spaces identified within the (digital) architecture of the platform.67 How these “spaces” are defined depends on the nature of the platform. For crowdworking platforms that require workers to have a certain level of specialization (e.g. PeoplePerHour or Fiverr, providing services in translation, graphic design, music production, programming, photography or digital marketing), the locus for the election of trade union delegates and workers’ representatives could perhaps be identified within the different typologies of the services offered in digital marketplaces. This approach would allow the representative body to build on the common skills and professional identity of the workers.68 Indeed, it could be argued that, by focusing on the sale of different types of products (or rather, of the labour used to realize a gig), these platforms are organized in what, in an offline world, would correspond to different departments, units, and ultimately, establishments. Depending on the platform’s size, geographical criteria based on the worker’s place of residence could also be applied to give voice to regional expressions of interests and demands.69
Defining a space for workers’ representation can be even more difficult in the case of crowdworking platforms focusing on microtasks where no specific skills or qualifications are required.70 These platforms operate by breaking jobs and projects into micro-activities, often lasting only a few minutes, which are then re-integrated by the platform to deliver a specific output.71 Within a single work session, workers perform simple but diverse tasks, making it nearly impossible to associate them with a particular segment or department of the platform.72
Without a discernible organizational link to space, even a virtual one, there might be an option to interpret the notion of “establishment” as encompassing the entire workforce providing labour for the digital marketplace. Alternatively, trade unions should be granted discretion to define an adequate locus for workers’ representation, designed to enable the effective exercise of information and consultation practices. An interesting forerunner example from the “times before” can be found in the 2002 Framework Agreement on Telework where the European social partners, in defining the conditions for teleworkers to set up bodies representing workers, specified that “the establishment to which the teleworker will be attached for the purpose of exercising his/her collective rights is specified from the outset”.73
The challenges of determining the space for the exercise of collective labour rights extend beyond the ambiguous notion of “establishment”. As mentioned earlier in this section, the EU acquis also links the exercise of information and consultation rights to the legal notion of “undertaking”. This notion not only relates to the dimensional reality of where such rights are exercised but also often coincides with the legal entity whose representatives act as counterparts to worker representatives in information and consultation practices.
Here, the challenge to the effective implementation of collective labour rights lies in the currently prevailing interpretation of the notion of “undertaking” within the domain of labour law. This interpretation still coincides with the concept of a legal entity established and recognized under company law: generally, any legal body, or even any natural person, carrying out a trade or business. This implies that the EU legal system does not dispose of a legal concept that captures the “economic galaxy” resulting from the extensive outsourcing practices enabled by digital labour platforms.74 It is not uncommon for conventional businesses to externalise non-core activities to platforms, which in turn increasingly subcontract the provision of the service to umbrella companies (as we see in the food or parcel delivery sectors) or to a team of self-employed workers managed by an individual contractor (as we see in online digital labour platforms). 75 A contradiction thus emerges: while EU law encompasses the micro reality of individual service providers without capital investment within the concept of “undertaking”,76 it fails to capture the full extent of the (decisively more influential) economic entities that heavily rely on outsourced labour to function and dominate the market.77 These powerful economic actors can therefore discreetly evade a substantial portion of the responsibilities normally associated with the employer,78 including information and consultation obligations.
We argue that the transformative impact of the platform economy calls for a refocusing on the bigger picture to ensure that the consequences of fissuring tendencies are adequately addressed through a purposive reinterpretation of the notion of “undertaking”. As Weil points out, while firms are deconstructed and formally downsized, leading corporate actors are continuing to exercise strict control over the productivity of satellite businesses through methods of monitoring and enforcement which, in the platform economy, are empowered by algorithmic management.79 Platforms have thus acquired increasingly pervasive labour market power, manifested in the ability to set wages on the basis of conventional pricing considerations.80
If no appropriate initiatives are taken, the extreme fragmentation caused by intricate subcontracting chains will undermine the effectiveness of information and consultation practices, even if the quantitative thresholds set by EU directives are met. These emerging labour market dynamics therefore require a reinterpretation effort to ensure that the scope of collective labour rights, as well as the identification and composition of the "managerial counterpart" in information and consultation practices, reflects the progressive centralization of economic and bargaining power in the hands of a few business actors.
There are several avenues that could be explored in this regard. Drawing on Hassel and Sieker’s conceptualization of the firm as a nexus of contracts, a reinterpreted notion of “undertaking” could encompass the economic and bargaining pressures manifested throughout the network of subcontracting governed by (or occurring through) digital labour platforms.81 Another approach, as suggested by Weil in his search for the “organizational glue” that defines the boundaries of the modern undertaking, is to look at the detailed supply chain standards and franchise agreements.82 Alternatively, another relevant indicator for locating the network of economic and contractual power could be to follow the algorithmic management thread used to allocate, direct and evaluate performances.83 Finally, inspiration could be derived from competition law where, in order to assess market distortions, the the European Commission and the Court of Justice have extended the notion of “undertaking” to include the “organic and functional links between firms”.84 Although a similar approach has yet to be consolidated in the area of labour law, a recent positive development occurred in the Ellinika case, where the Court acknowledged that, if one undertaking depends on the economic decisions of another, its autonomy as a separate legal entity may not be genuine.85
While these are only a few of the possibilities available, and we are aware that some may be more feasible than others, we believe that it is crucial to persuade the policymaking institutions to undertake similar innovative reinterpretation efforts. This re-conceptualization is essential to ensure that the scope and effectiveness of collective labour rights matches the actual labour market power wielded by powerful employers in the expanding platform economy.
On-call, waiting and travel time (unpaid)
One foundational rule of platform work provides that workers are “engaged” (and remunerated) by the platform only when they accept an assignment having logged in to the app. The implications of this model are multifarious.86 A key consequence of such “artificial separation of the concept of performance from the concept of remuneration” is the inordinate dominance of unpaid on-call, waiting and travel time and the compounding necessity to work long hours to earn decent wages.87 The amount of unpaid time spent waiting for or securing tasks is similar to the time spent completing those tasks.88 Online workers spend around one-third of their effective working time unpaid.89 Hyperflexible arrangements are also prevalent in traditional fields like care, cleaning, customer support, higher education and logistics, reflecting the same kind of temporal division that is seen in the gig economy.
The Working Time Directive (WTD)90 can offer some important guardrails but it presents limitations that need to be addressed. It defines working time as “any period during which the worker is working, at the employer’s disposal and carrying out [her] activity or duties, in accordance with national laws and/or practice” (Article 2(1)).91 The notion has been expansively and purposively interpreted by the CJEU as a “unitary one” embracing all those expanses of time units, including “non-productive” ones, when workers work or even make themselves available to an employer’s call to provide appropriate services at short notice.92 The WTD’s main provisions are minimum weekly and daily rest periods, minimum daily breaks and annual leave, and maximum weekly and night working time.93 Space limitations do not permit us to dwell on the coveted discussion of the choice of the directive’s legal basis. Suffice it to remark that opting for a foundation related to working conditions rather than health and safety (Article 153(2) TFEU) would have allowed the inclusion of a larger set of issues related to working time, namely “short or variable working hours or workers” as well as control over work schedules and arguably remuneration.94 The enforcement of working time provisions does not fully address the risks and precariousness faced by non-standard forms of work, nor does it curb their use both as regards the way in which the provisions are framed and as regards the prevalence of opt-out clauses.
There are three main critical aspects that are hardly captured by the WTD: (a) on-call time; (b) waiting time; and (c) travel time. These issues are ubiquitous in platform work.95
On-call (or stand-by) time was the crux of several rulings handed down by the CJEU, well before the advent of platform work. Significantly, when stand-by time is spent at the workplace, broadly understood as “any place where the worker is required to exercise an activity at the employer’s instructions, including where that place is not the place where he or she usually carries out his or her professional duties”,96 even if the worker is not carrying out “productive” activities in the interest of the employer, this does not automatically entail such time counting as rest. An aspect that deserves attention is the trend towards overcoming the centrality of the so-called “location test”,97 once considered a key determinant for distinguishing between working time and rest periods in the case of on-call workers. A leading criterion used by the CJEU, which conceives time in a qualitative rather than simply a quantitative way,98 in identifying these time units is the more or less significant restriction of a worker’s opportunities to engage in other activities or to rest.99 Being at a location different from the company premises while “at the employer’s disposal” amounts to working time if the workers’ freedom to engage in other personal or social activities is significantly constrained.100 This is the case for many platform workers who are dissuaded from dedicating themselves to other activities, including recreational ones, as they need to be ready to accept a call and, as a result, are constantly checking on the app’s notifications. The open clause of “constraints” has been complemented by references to their “objective” and “significant” nature in a recent CJEU ruling which partially displaces the solid, binary demarcation defined in previous decisions,101 paving the way for alternative interpretations justified in the light of organizational needs. In the absence of such constraints, only the time linked to the provision of work carried out during that period constitutes “working time”. Moreover, the CJEU has held that the WTD does not preclude a national law, collective agreement or a decision of an employer treating those periods during which work is materially carried out and those during which no actual work is accomplished (both to be considered working time) in different ways for the purposes of remuneration.102
A stark mismatch emerges between the business model adopted by most platforms, geared towards consideration for the sole time spent logged or engaged in the task commissioned,103 and how time spent in the interest of the employer is considered in the CJEU jurisprudence. That a worker, after logging in, is awaiting a call or spending time skimming through tenders launched by potential clients to select the best-paying options far from qualifies as rest time and cannot be seen as a “third category” of time between working time and rest.
The very architectural makeup of the platforms’ business model is based on the “implied” permanent availability and oversupply of workers. Coupled with piece rate (output-related) income arrangements and gamification tools, this model is the cause of squeezed pay levels. It goes without saying that, if workers are so constrained that they cannot freely manage their time and pursue their own interests even when their professional services are not required, this time must be classified as working time. Asserting that platform workers could be completing other side activities to the benefit of competing platforms when their services are not needed would result in exacerbating the paradox of interpreting the notion of working time in a way that endorses the erratic nature of platform arrangements, thereby jeopardizing the overarching protective purpose of the WTD. Admittedly, waiting time represents a Hamlet dilemma for offline and online platform workers. By using minimum rest time, they may witness a decrease in their income or be penalized by algorithms fuelled by metrics such as constancy and availability. By committing to complete more gigs for competitors, they intensify their psychosocial risks and reduce their chances of being reclassified should they lodge a complaint in jurisdictions where the presence of non-exclusivity and substitution clauses is used to defeat the “personal work” factor.104
What about travel time? Offline platform workers constantly experience this condition: they must reach a “hotspot” where they can log in to serve a certain urban area with no “guarantee of obtaining any paid activity after investing time in travelling to work”.105 In a ruling considering the journey completed from home to the first client and from the last client to own home by a peripatetic worker without fixed places of work,106 the CJEU has stated that even travel time counts as working time insofar as the worker is required to be physically present at a place determined by the employer and to be available to carry out work. In the absence of a fixed or habitual place of work, the journey is considered technically indispensable to rendering the service.107 While the CJEU’s adaptive posture is promising from the perspective of platform workers, it must be added that, for the purposes of the WTD, qualifying travel time as working time does not entail that it must be remunerated,108 a prerogative that rests entirely with the platform employer.109
Another general principle asserted in CJEU case law is that employers must be required to set up an objective, reliable and accessible system enabling the time worked each day by each worker to be measured.110 It is striking that, in highly tracked and constantly monitored environments, there is very little information on the number of hours worked, a situation that further exacerbates information asymmetries to the advantage of platform companies. On top of this, one major drawback of the WTD is that, despite the progressively expansive interpretation of its material scope, it is deafeningly silent on how to remunerate the time spent working. This black hole in the regulation of time also emerged in the context of one of the CJEU’s very first reasoned orders seeking to deal with a company organization that squarely fitted that of a major platform operator.111 No answer was provided by the CJEU to the question of how to measure the working time of workers who are allowed to “multi-home”; namely, to be available on multiple competing platforms at the same time. The feasibility of nominal multitasking cannot be doubted enough. Despite this, no guidance has been offered as regards the economic treatment of patchwork engagements, something which is destined to have a negative impact on health and safety conditions.
Another weakness is that the WTD simply sets upper limits, in line with the spirit of the very first ILO Convention 1919 (No. 1). However, working time is understood in a narrow sense, not as an integral component of working conditions, but rather as one of the means to promote safety and health. This notion may result in unintended consequences. Platform workers can see the economic convenience of not being covered by working time provisions and self-exploit to increase their remuneration. As a vicious circle, this perverse incentive potentially pushes many of them into bogus status classification, making the category of self-employment more attractive with severe individual and societal effects. All in all, the WTD serves as a key illustration of the cases where it falls short of achieving its “humanization” objectives within the realm of platform work. This is where the arrangement of work should be adapted to meet the needs of workers as human beings, but the WTD does not come close to succeeding in enforcing such an outcome.
Opaque and unpredictable schedules
One of the main sources of temporal precariousness for casual workers is the unpredictability of future workloads, precipitated by the employer’s intense degree of control over the whole labour process.112 The combination of hyperflexible business models and the strategy of oversupply makes workers prey to “hours famine”.113 Since the internal architecture is “designed to keep (or have the effect of keeping) the worker hungry for the next shift”,114 current working time rules risk being a blunt weapon. In fact, temporal precariousness is particularly affecting already disadvantaged and vulnerable workers.115 Up until recently, little could be found in the social acquis to make sure that minimum working hours were guaranteed to workers to avoid time drainage and foster programmability, clarity and balance between working and private spheres.
Meagre results have been achieved by mobilizing the principles of non-discrimination before the CJEU. In a case brought by an “on-demand” worker in the retail sector (hired on a zero hour basis, with the right to refuse assignments), the Court ruled that there was no “comparator” in the enterprise for the purposes of the assessment mandated by the framework agreement on part-time work.116 The justification was that all full-time workers had fixed hours and were required to accept work without the option to refuse it, as stipulated by their contracts.117 Analogue reasoning would be fatal in contexts where the wholesale business model relied on “lean workforces”, operating almost exclusively through self-employed workers. Such a company structure would make any benchmarking activity complex and burdensome. A narrow interpretation of the notion of comparator would thus displace the application of the equality principle, which is one of the possible remedies to the dangerous unpredictability of erratic arrangements.118 Not surprisingly, the “safe conduct” opened by the lack of comparable standard workers represents yet another incentive not to invest in building more solid business models.
Temporal discontinuity rather than genuine autonomy has been used in court to rebut arguments that platform workers are in fact employees. Yet, it is precisely the subjection to highly casualized schedules, also used by platforms to discipline workers,119 that should be read as a strong indicator of subordination, understood as technical subjection to a contractual party who discretionally exercises command and control prerogatives.120 Organizational unpredictability and contractual impermanence by design cannot result in detriment to workers, nor can these issues be weaponized in terms of a waiver of fundamental rights.121 Here lies another incongruity of the current legal framework, which is poorly equipped to deal with non-standard arrangements.
Some compelling responses to the shortcomings presented above could come from the implementation of Principle 5 (the right to fair and equal treatment regarding working conditions and the prevention of employment relationships that lead to precarious working conditions) and Principle 7 (the right to receive information about employment conditions and protection in the case of dismissal) of the European Pillar of Social Rights (EPSR).122 By giving partial implementation to those principles, the provisions fleshed out in the Directive on Transparent and Predictable Working Conditions represent a step forward against the excess of unidirectional flexibility that “has rendered working relationships unstable and increasingly unpredictable”.123
The two key goals of this directive are to ensure workers’ “time sovereignty” over the labour process and to avoid involuntary underemployment. Thanks to its legal basis (Article 153(2) TFEU and point (b) of Article 153(1) TFEU), the TPWCD relies on a broad notion of programmability encompassing issues such as a guaranteed number of paid hours, payments for additional work, reference hours and days when the worker may be required to work, minimum notice periods and deadlines to revoke availability or, on the part of the employer, to cancel an assignment.124 It is safe to claim that the TPWCD integrates and advances the scheme set in place by the WTD, tackling some salient issues that had been unaddressed and offering reliable safeguards for all workers with short, insecure or variable work hours. The TPWCD’s rationale is that the precondition for improving working conditions and reducing precariousness is to provide workers with accurate information. It represents a regulatory paradigm shift as it looks at working conditions with a more pragmatic and comprehensive approach. On the one hand, it updates and expands the set of information to share and strengthen deadlines; on the other, it expressly covers atypical and casual workers, calibrating its target on the ability to (pre)determine schedules and organize private lives.125
At ILO level, a comparable instrument can be found in the ILO Domestic Work Convention 2011 (No. 189), which provides domestic workers with the right to be “informed of their terms and conditions of employment in an appropriate, verifiable and easily understandable manner and preferably, where possible, through written contracts in accordance with national laws, regulations or collective agreements” (Article 6). Such information is with reference to, among others, the starting date and duration (in the case of temporary contracts), remuneration, method of calculation and periodicity of payments, normal hours of work, paid annual leave, daily and weekly rest periods.126
Returning to the crux of the matter, every worker “who has an employment contract or employment relationship as defined by the law, collective agreements or practice in force in each Member State with consideration to the case-law of the Court of Justice” falls under the scope of the TPWCD (Article 1). As long as they fulfil the (ample and purposive) criteria set by the CJEU, domestic, on-demand, intermittent, voucher-based and platform workers, trainees and apprentices fall within the scope of the TPWCD and are entitled to a set of minimum rights127, except those whose predetermined and actual working is equal to or less than an average of three hours per week in a reference period of four consecutive weeks (this provision could have a negative impact on those platform workers who experience periods of peaks and troughs). However, this “hybrid” worker definition, departing from the “static” one without attaining the greater ambition that was proposed in other versions of the directive,128 paves the way for a stronger role for the CJEU which could be asked to conduct a scoping exercise on a case-by-case basis. Since the TPWCD embeds the principle of the prevalence of substance over form, it could be instrumental in bringing more workers than those who are protected by national labour law within the scope of the EU social acquis.129 National courts have started to be sensitive to this opportunity in the context of status litigation.
The “second generation” of atypical employment is the ideal beneficiary of the protective provisions provided for by the TPWCD. The directive lays down information rights about the organization and the number of guaranteed paid hours. Workers must be informed on how they will be paid for the additional hours worked, when exactly their assignments will start and within what timeframe a call can be cancelled (Article 4). When the work organization is “entirely or mostly unpredictable”, the directive’s provisions design a clear availability period. Two conditions must be fulfilled: (a) work is demanded within predetermined reference hours and days; and (b) workers are informed within a reasonable notice period. If one or both conditions are not met, workers can refuse to show up without suffering adverse consequences. In addition, they must be compensated if previously agreed slots are cancelled (Article 10). According to Article 12, workers also have the right to request to switch to a new form of employment with more predictable and secure working conditions (and employers have an obligation to reply). However, this does not mean that any obligation to “convert” the contract arises from the TPWCD.
Minimum requirements are designed to prevent abuses in volatile segments of the economy. Simultaneous engagements with other employers are allowed; that is, the worker will not be treated adversely and incompatibility clauses must be limited (Article 9). The purpose of this prohibition is praiseworthy: it aims at preventing hold-up situations where workers are tied to a single employer and denied the opportunity to hold parallel positions with other employers. However, these concessions are contentious.130 On the one hand, several platforms insert non-exclusivity clauses – which often remain confined to the letter of the contracts – to reinforce their arguments in respect of workers’ alleged autonomy. On the other, legitimizing the practice of taking multiple jobs leads to further work intensification. Workers agree to work on multiple tasks, assignments and projects because they are left with no choice if they are to earn a meaningful level of pay. Workers’ (in)voluntary choice should not find technical encouragement and tolerance in the law.
The directive envisages a set of rules aimed at preventing and fighting abuses. These may include limitations on the use and duration of on-demand or similar contracts – in line with similar provisions included in the framework agreement on fixed-term work131 – or the introduction of a rebuttable presumption regarding the existence of an employment contract or employment relationship with a minimum amount of paid hours calculated on the basis of the average hours worked during a given period (Article 11).132 Although these provisions purport to re-standardize casual work, they replicate “some of the paradoxes of current EU employment policy”,133 as it results in trading a reasonable level of predictability for employees with large margins of flexibility offered to employers. This compromise approach risks resulting in an excessive normalization and promotion of atypical work, depicted as collateral damage stemming from the need to ensure labour market adaptability during uncertain times.
Pervasive monitoring, arbitrary ratings and capricious management
When it comes to attributing responsibility for degraded working conditions in platform work, the finger should be pointed at organizational structures that rely on a blend of algorithmic management (AM) and human supervision.134 Platforms are unanimously identified as the birthplace of pervasive monitoring and automated decision-making systems.135 While courts have been enmeshed in litigation campaigns concerning workers’ employment status, the riskiest features of platform work have become a widespread model.136 Gradually, conventional jobs in ordinary industries have been exposed to human resource practices that are completed or at least supported by digital instruments and software.137 Undeniably, remote work prompted by the pandemic has contributed to normalizing the use of monitoring devices and productivity-tracking tools.
The opaqueness of AM is strongly correlated with the difficulties of ensuring the proper classification of platform workers: data-driven practices both disguise and intensify managerial prerogative. Judges called on to assess the factual circumstances of the relationship have struggled to understand the intricacies of these new organizational patterns for exercising command and control and to match them with classical legal and interpretative categories. Only recently have they become familiar with the overarching implications of AM systems on status and working conditions.
This accrued knowledge, however, is not fully leveraged in the draft directive on platform work. Although control over work performance is the trigger for the rebuttable presumption of employment (Article 4(2)), and consideration for the use of algorithms is indicated in connection with implementing the primacy of facts principle (Article 3(2)), no explicit reference to AM systems is made in the list of five indicators supporting the process under which control can be established. Almost all criteria, such as determining the remuneration, issuing instructions on how to behave, supervising the execution of performance and constraining the ability to self-organize one’s work, can be generally completed by AM systems. The lack of such a mention shows a far-from-perfect integration between chapters II and III of the proposed platform work directive.
Platform workers cannot clearly understand the consequences of their conduct and find themselves dealing with capricious agents. In uncertain situations, they suffer from “algorithmic insecurity” (the vulnerability and fear stemming from platforms’ use of customer-generated ratings to score workers and algorithms to amplify the repercussions of those grades).138 Workers must guess the metrics considered by the internal ranking system by constantly adhering to an implicit standard of conduct that reduces their self-determination.139 They have little opportunity, if any, to influence the logic of AM systems besides being aware of the constant monitoring and their exposure to arbitrary sanctions140 which affect their career prospects and potential earnings. Platform workers end up overworked, disempowered, stressed and, ultimately, less satisfied with their jobs.141 Concomitantly, the absence of employment status, exacerbated by the scarce application of the canonical substantive and procedural limits to employers’ powers, renders the subjection to AM even more detrimental and hazardous.142 Workers are also discouraged or prevented from exercising collective rights and have limited access to fundamental rights to establish and join organizations that promote social dialogue.143 In addition, several investigations and rulings have documented how internal scores are used to terminate workers’ accounts when they fall below a certain threshold owing to causes independent of workers’ will and in defiance of international standards.144
The proposed platform work directive is the first EU instrument that directly addresses data-driven organizational practices. It is praiseworthy that provisions ensuring fairness, transparency and accountability with regard to algorithmic management in the platform work context cover all workers regardless of their contractual designation and employment status. The text presented by the Commission in December 2021 extended the ambit of application of the chapter on AM to “persons performing platform work […] who do not have an employment relationship, i.e. the genuine self-employed”. Yet, these safeguards do not go beyond the boundaries of the gig economy, thus limiting the draft’s ambitious impact. Legions of workers in conventional sectors are now managed by digital systems that partially or wholly assume managerial functions. The information and access rights enshrined in the proposed platform work directive are not available to them unless they are organized by a “digital labour platform” as defined under Article 2 of the proposed platform work directive (with emphasis on the role of “digital electronic means, such as a website or a mobile application”).145 A landmark achievement could turn out to be a Pyrrhic victory.
The new rules on AM aim to reduce the information asymmetries between platforms and workers.146 They cement universal principles of procedural fairness and the transparency of managerial decisions. Significantly, they could improve working conditions by addressing the accountability deficit documented in a myriad of studies,147 besides facilitating access to the evidence to be used in court to demonstrate the existence of the robust “control” being exercised by platform businesses.
The platform work directive targets the “automated monitoring systems” used to monitor, supervise or evaluate platform workers and the “automated decision-making systems” implemented to “[m]ake or support decisions that significantly affect platform workers’ working conditions”. According to Article 6, platform workers must be informed about the existence of or the intention to adopt these practices, the categories of action monitored, the decisions made or supported by such systems, the main parameters taken into account and their relative weight, and the grounds for impactful decisions such as restriction, suspension and termination of the workers’ account. Importantly, the relevant information must be shared with worker representatives or authorities upon their request. Genuine self-employed workers are excluded from this collective information and consultation right and, in principle, can resort to Article 5 of the P2B Regulation148 according to which they are entitled to learn from terms and conditions “the main parameters determining the rankings and the reasons for the relative importance of those main parameters as opposed to other parameters”.
We have already pointed to the contradiction of designing a sectoral model of safeguards for a trend that is rapidly sweeping through the entire labour market, leaving vulnerable workers particularly exposed to its dangers. To this must be added the ambivalent relationship between the platform work directive and the GDPR. The former is significantly less detailed, yet slightly broader in material scope than the latter. In line with Article 22 GDPR (the right to obtain human intervention, to express a point of view and to contest a decision), the platform work directive’s redress mechanism includes a human assessment of significant decisions on the request of the affected person and a review of the decision if the explanation does not prove compelling or workers feel that their rights have been hindered. The platform work directive provisions partially address the pitfall of a narrow emphasis on solely automated decision-making (Article 22 GDPR), which has also emerged in the case law,149 by including decisions that are merely supported by automated decision-making systems. The platform work directive also introduces “the right to obtain an explanation” for decisions such as access to tasks, earnings, working time, promotion, status and the restriction, suspension or termination of their accounts (Article 8). Despite this, automated monitoring and management, whose adoption is never questioned in the platform work directive, are considered inherent (or perhaps inevitable) functions of the platform business model. Nor do information and consultation rights equate to the strongest codetermination models that are in force in many EU countries, where worker representatives must be meaningfully involved in the process of technology adoption.150
There is, however, a concomitant EU regulatory initiative, the so-called AI Act,151 that casts doubt on the willingness to establish a barrier against the undue proliferation of toxic automated decision-making systems. AI systems adopted in the context of “employment, workers management and access to self-employment” (point 4, Annex III) are identified as posing “high risk to the health and safety or fundamental rights of persons”. They must undergo a conformity assessment procedure prior to entering the market or being put into service (Chapter 2). This procedure involves establishing, implementing, documenting and maintaining a risk management system throughout the life cycle of the AI system.152
The resulting patchwork of instruments opens a “constitutional” minefield at supranational level. Given its legal basis is aimed at promoting maximum harmonization, there is a growing fear that the AI Act could upset domestic rules on the adoption and deployment of workplace technologies, including the newly proposed rules in the context of platform work. Article 20 of the proposed directive on platform work empowers Member States to apply or introduce laws, regulations, administrative provisions or collective agreements which are more favourable to platform workers, in line with Article 88 GDPR on the possibility of adopting “more specific rules” to protect employees’ personal data at work. Should the AI Act be approved in its current formulation, more stringent provisions could be considered exorbitant barriers to the free provision of AI-related services in the EU. One actionable solution would be the removal of AI systems used in working environments from the AI Act, to avoid tensions with current national monitoring and data protection rules.
The limitations of the inflated EU framework and the uncertainty raised by partially overlapping models make the picture highly unstable, and rights scarcely actionable. Due to their idiosyncratic scope of application, the AM provisions in the platform work directive could be easily circumvented and, moreover, they have scarce bite. At national level, the situation is even more complex given the contentious coexistence between domestic laws implementing the GDPR regarding personal data collection and processing,153 the labour-related frameworks governing the adoption and deployment of technology at work and the enforcement role being scattered among labour tribunals and data protection authorities.
Given the high stakes at play, what is chiefly needed is a comprehensive supranational, or at least European, instrument that regulates workplace data protection154 and the use of technologies to perform monitoring and decision-making functions at work. An EU-wide model would reduce the costs connected with compliance and prevent regulatory arbitrage. Borrowing the technique adopted for identifying the five-factor test to trigger the presumption of employment, a selection of domestic traditions and rules could be embedded and harmonized at EU level, minimizing the risk of rejection.
Workplace technologies must be introduced upon conducting a multi-stakeholder process that involves those people who are affected by such tools, not only the providers and the deployers. In certain critical areas, decisions ought not to be left to AM.155 The unique traits of professional relationships, including unequal access to information and pre-established imbalances in contractual power, instead demand more specialized regulations.
Considering the significant risks and the difficulties presented by AM, a more focused approach must be developed. There are several good reasons for doing so, among them the opportunity to co-design, update and fine-tune those data-driven managerial practices that are not dysfunctional or prone to be gamed. Trade unions and workers’ representatives need to wield the institutional power to codetermine the establishment of technology at work and to negotiate the action, decisions and metrics that are monitored or factored in by AM tools. There are numerous pioneering experiences which demonstrate that a fruitful social dialogue on technology is beneficial for both workers and companies.156 Harnessing the benign potential of digital transformation is an opportunity not to be sacrificed on the altar of techno-determinism.
Conclusion
This paper offers a critical analysis of both the proposals for an EU directive on the working conditions of platform workers and other instruments currently in the legislative pipeline that, if adopted, would be the first supranational attempt to regulate the now-established phenomenon of platform work. Our goal was to match the proposed platform work directive with the EU and international “social acquis”.
The paper begins by approaching platform work as an expression of the broader phenomenon of precarious and non-standard work. In fact, the vast majority of the challenges and “protection gaps” we have identified also apply to an increasingly wide range of workers who share similar characteristics with platform workers. Casual work formats and remote working arrangements are two important examples. We contend that these phenomena are mutually reinforcing and we hope that, by highlighting the main advantages and limitations of the EU legal framework in regulating working conditions in the platform economy, our work can contribute to shaping research and, most urgently, the policy agendas aiming at the establishment of a fairer and more worker-centred future of work.
Rather than analyse the directive’s proposals in a systematic way (a type of enquiry already carried out by several other experts)157 the paper has explored how some of its provisions (for example those defining its scope, those introducing a legal presumption of status, those defining the concept of workspace and workplace, and therefore the concept of employing entity, and the rules on surveillance and monitoring) interact with those contained in other EU instruments (the TPWCD, WTD and GDPR, to name the most significant), generating both some overlaps but also, and crucially, some “underlaps” and protective lacunae.
Overall, there is no denying that the proposed directive, if adopted, would introduce a number of new protective devices that would, if properly implemented, determine tangible improvements for platform workers. In some respects, this instrument would go as far as providing a blueprint of rights and protections, for instance in the areas of digital surveillance and algorithmic monitoring, that, if suitably adapted, could benefit all workers and not just those working through platforms.
But it is also clear that the proposed directive is at risk of falling short of resolving the (constantly developing) concerns that algorithmic management and platform work generate for workers. The more specific reform suggestions explored in the previous sections aside, this report suggests that – for this directive to achieve its protective objectives – it would be important to adapt some other existing instruments to the regulatory challenges posed by platform work and non-standard forms of employment.158
The paper also suggests that the proposed directive on platform work should take it upon itself to provide national implementing authorities and national and European courts with a series of guiding principles assisting them with coordinating and reconciling a number of scattered sources contained in multiple instruments that were not designed to “speak to each other”. In this respect, the seemingly haphazard references to instruments such as the WTD, the TPWCD, the directive on information and consultation and the GDPR fail to “connect the dots” and provide a coherent ecosystem of rules regulating platform work in its entirety. There is also a greater need to ensure that an instrument such as this is genuinely “future proofed”, at least in terms of the vexed question of the typology of workers (and the concept of work) to which it should apply. In that regard, some specific suggestions have been advanced in this paper, suggestions that are – we believe – fully in line with the dicta of the Universal Labour Guarantee proposed by the ILO Global Commission on the Future of Work which states that, “All workers, regardless of their contractual arrangement or employment status, should enjoy fundamental workers’ rights…”.159
References
Abraha, Haleform H. 2022. “A pragmatic compromise? The role of Article 88 GDPR in upholding privacy in the workplace”. International Data Privacy Law 12 (4): 276-296.
Abraha, Halefom H. 2023. “C-34/21 - Hauptpersonalrat v Wiesbaden: Article 88 GDPR and the Interplay between EU and Member State Employee Data Protection Rules”. Available at SSRN: https://ssrn.com/abstract=4446107. Forthcoming in Modern Law Review.
Adams, Zoe and Nicola Countouris. 2019. “Heteromated labour in 21st century computing capitalism: A critical conceptualisation of ‘work’”. Proceedings of the 6th conference of the ILO ‘Regulating for Decent Work’. Geneva: International Labour Office.
Adams-Prassl, Jeremias. 2022. “Regulating algorithms at work: Lessons for a ‘European approach to artificial intelligence’”. European Labour Law Journal 13 (1): 30-50.
Aimo, Mariapaola. 2020. “Lavoro a tempo, on demand, ‘imprevedibile’: alla ricerca di una ‘ragionevole flessibilità’ del lavoro non-standard”. Working Papers – Centro studi di diritto del lavoro europeo ‘Massimo d'Antona’.
Aloisi, Antonio. 2020. “‘Time is Running Out’. The Yodel Order and Its Implications for Platform Work in the EU". Italian Labour Law e-Journal 13 (2): 67-87.
Aloisi, Antonio. 2022. “Platform Work in Europe: Lesson Learned, Legal Developments and Challenges Ahead". European Labour Law Journal 13 (1): 1-26.
Aloisi, Antonio. (2023). “J.K. v TP S.A. and the ‘universal’ scope of EU antidiscrimination law at work: a paradigm shift?”. Industrial Law Journal.
Aloisi, Antonio and Valerio De Stefano. 2022. Your Boss Is an Algorithm: Artificial Intelligence, Platform Work and Labour. Oxford: Hart Publishing.
Aloisi Antonio and Elena Gramano. 2019. “Artificial intelligence is watching you at work. Digital surveillance, employee monitoring and regulatory issues in the EU context”. Comparative Labor Law & Policy Journal 41 (1): 95-121.
Altenried, Moritz. 2020. “The platform as factory: crowdwork and the hidden labour behind artificial intelligence”. Capital & Class 44 (2): 145-158.
Atkinson, Joe and Hitesh Dhorajiwala. 2019. “IWGB v RooFoods: Status, rights and substitution”. Industrial Law Journal 48 (2): 278-295.
Ball, Kirstie. 2010. “Workplace Surveillance: An Overview”. Labor History 51 (1): 87-106.
Bednarowicz, Bartłomiej. 2019. “Delivering on the European Pillar of Social Rights: The New Directive on Transparent and Predictable Working Conditions in the European Union”. Industrial Law Journal 48 (4): 604-623.
Bednarowicz, Bartłomiej. 2020. “The tale of transparent and predictable working conditions intertwined with work-life balance: Assessing the impact of the new social policy directives on decent working conditions and social protection”. European Journal of Social Security 22 (4): 421-433.
Bell, Mark. 2012. “Between Flexicurity and Fundamental Social Rights: the EU Directives on Atypical Work”. European Law Review 37 (1): 31-48.
Bell, Mark. 2019. “EU Equality Law and Precarious Work”. In Uladzislau Belavusau and Kristin Henrard, eds. EU Anti-Discrimination Law Beyond Gender. Oxford: Hart Publishing, pp. 75-94.
Bérastégui, Pierre. 2021. Exposure to psychosocial risk factors in the gig economy - A systematic review. Brussels: ETUI.
Berg, Janine, Marianne Furrer, Ellie Harmon, Uma Rani and M. Six Silberman. 2018. Digital labour platforms and the future of work: Towards decent work in the online world. Geneva: International Labour Office.
Blanpain, Roger. 2002. Involvement of Employees in the European Union: European Works Councils, The European Company Statute and Information and Consultation Rights. Deventer: Kluwer Law International.
Boewe, Jörn, Tina Morgenroth and Johannes Schulten. 2021. “Amazons Letzte Meile”. Thüringen: Rosa-Luxembourg-Stiftung.
Bogg, Alan. 2016. “The regulation of working time in Europe”. In Alan Bogg, Cathryn Costello and C.L. Davies, eds. Research Handbook on EU Labour Law. Cheltenham, UK: Edward Elgar Publishing.
Bogg, Alan. 2019. “Taken for a ride: workers in the gig economy”. Law Quarterly Review 135: 219-226.
Cefaliello, Aude, and Cristina Inversi. 2022. “The impact of gig economy on occupational health & safety: just an occupation hazard?” in Valerio De Stefano, Ilda Durri, Charalampos Stylogiannis and Mathias Wouters, eds. A Research Agenda for the Gig Economy and Society, Cheltenham, UK: Edward Elgar, pp. 33-52.
Cefaliello, Aude and Miriam Kullman. 2022. “The draft Artificial Intelligence Act (AI Act): Offering false security to undermine fundamental workers’ rights”. European Labour Law Journal 13 (4): 542-562.
Collins, Philippa and Joe Atkinson. 2022. “Worker voice and algorithmic management in post-Brexit Britain”. Transfer: European Review of Labour and Research 29 (1): 37-52.
Countouris, Nicola. 2016. “EU Law and the Regulation of ‘Atypical’ Work”. In Alan Bogg, Cathryn Costello and C.L. Davies, eds. Research Handbook on EU Labour Law. Cheltenham, UK: Edward Elgar Publishing.
Countouris, Nicola. 2019. Defining and regulating work relations for the future of work. February. Geneva: International Labour Office.
Countouris, Nicola and Valerio De Stefano. 2019. New trade union strategies for new forms of employment. Brussels: ETUC.
Countouris, Nicola and Valerio De Stefano. 2023. “Out of sight, out of mind? Remote work and contractual distancing”. In Nicola Countouris, Valerio De Stefano, Agnieszka Piasna and Silvia Rainone, eds. The Future of Remote Work, Brussels: ETUI.
Daskalova, Victoria. 2022. “Rethinking Collective Bargaining for the Self-Employed: European Commission publishes Guidelines on Exemption and non-Enforcement”. Kluwer Competition Law Blog. 7 October. https://competitionlawblog.kluwercompetitionlaw.com/2022/10/07/rethinking-collective-bargaining-for-the-self-employed-european-commission-publishes-guidelines-on-exemption-and-non-enforcement/
de Groen, Willem Pieter, Zachary Kilhoffer, Leonie Westhoff, Doina Postica and Farzaneh Shamsfakhr. 2021. “Digital Labour Platforms in the EU: Mapping and Business Models”. Brussels: Centre for European Policy Studies Commission, Brussels.
De Stefano, Valerio. 2016a. “Casual work beyond casual work in the EU: The underground casualisation of the European workforce – and what to do about it”. European Labour Law Journal 7 (3): 421-441.
De Stefano, Valerio. 2016b. “The Rise of the ‘Just-in-Time Workforce’: on demand work, crowdwork and labour protection in the ‘gig-economy’”. Working Paper. Geneva: International Labour Office.
De Stefano, Valerio. 2021. “Not as Simple as It Seems: The ILO and the Personal Scope of International Labour Standards”. International Labour Review 160 (3): 387-406.
De Stefano, Valerio. 2022. “The EU Commission’s proposal for a Directive on Platform Work: an overview”. Italian Labour Law e-journal 15 (1): 1-11.
Donini, Annamaria. 2019. “Il luogo per l’esercizio dei diritti sindacali: l’unità produttiva nell’impresa frammentata”. Labour and Law Issues 5 (2): 99-114.
Drahokoupil, Jan and Kurt Vandaele, eds. 2021. A modern guide to labour and the platform economy. Cheltenham, UK: Edward Elgar.
Dube, Arindrajit, Jeff Jacobs, Suresh Naidu and Siddharth Suri. 2018. “Monopsony in Online Labour Markets”, NBER Working Paper 24416. Cambridge, Mass: National Bank of Economic Research.
Duch-Brown, Néstor, Estrella Gomez-Herrera, Frank Mueller-Langer and Songül Tolan. 2022. Market Power and Artificial Intelligence Work on Online Labour Markets. Joint Research Centre Digital Economy Working Paper 2021-10. Seville: European Commission.
Dukes, Ruth. 2020. “Regulating Gigs”. Modern Law Review 83 (1): 217-228.
Dukes, Ruth. 2022. “On Demand Work as a Legal Framework to Understand the Gig Economy”. In Valerio De Stefano, Ilda Durri, Charalampos Stylogiannis and Mathias Wouters, eds. A Research Agenda for the Gig Economy and Society. Cheltenham, UK: Edward Elgar, pp. 133-148.
Eurofound. 2020. New forms of employment: 2020 update. Luxembourg: Publications Office of the European Union.
European Commission. 1975. Green paper on employee participation and company structure of the European Community. COM (75) 570 final.
European Commission. 2017. Impact Assessment Accompanying the Proposal for a Directive of the European Parliament and of the Council on Transparent and Predictable Working Conditions in the European Union, SWD (2017) 478 final.
European Commission. 2021a. Proposal for a Directive of the European Parliament and of the Council on Improving Working Conditions in Platform Work. COM (2021) 762 final, Brussels.
European Commission. 2021b. Better Working Conditions for a Stronger Social Europe: Harnessing the Full Benefits of Digitalisation for the Future of Work, COM (2021)761 final, Brussels.
European Commission. 2021c. Proposal for a Regulation of the European Parliament and of the Council laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and amending certain Union legislative acts, COM (2021) 206 final, Brussels.
European Commission. 2022. Guidelines on the application of Union competition law to collective agreements regarding the working conditions of solo self-employed persons, 2022/C 374/02, Brussels.
European Parliament. 2022. Report on the proposal for a directive on improving working conditions in platform work. A9-0301/2022. https://www.europarl.europa.eu/doceo/document/A-9-2022-0301_EN.html.
Ferrante, Vincenzo. 2019. “Between health and salary: The incomplete regulation of working time in European law”. European Labour Law Journal 10 (4): 370-385.
Fredman, Sandra, Darcy du Toit, Mark Graham, Aradhana Cherupara Vadekkethil, Gautam Bhatia and Alessio Bertolini. 2021. “International Regulation of Platform Labor: A Proposal for Action”. Weizenbaum Journal of the Digital Society 1 (1).
Gawer, Annabelle and Nick Srnicek. 201. “Online platforms: economic and societal effects”, Panel for the Future of Science and Technology. PE. 656.336. Brussels: European Parliamentary Research Service.
Georgiou, Despoina. 2022a. “‘Business Risk-Assumption’ as a Criterion for the Determination of EU Employment Status: A Critical Evaluation”. Industrial Law Journal 51 (1): 109-137.
Georgiou, Despoina. 2022b. “The new EU Directive on Transparent and Predictable Working Conditions in the context of new forms of employment”. European Journal of Industrial Relations 28 (2): 193-210.
Gellert, Raphael, Marvin van Bekkum and Frederik Zuiderveen Borgesius. 2021. “The Ola & Uber judgments: For the first time a court recognises a GDPR right to an explanation for algorithmic decision-making”. EU Law Analysis, 28 April. https://bit.ly/3uTR1Xt.
Gramano, Elena. 2022. “Stand-by time through the Court of Justice’s lens”. European Labour Law Journal 13 (4): 577-584.
Hassel, Anke and Felix Sieker. 2022. “The Platform Effect: How Amazon Changed Work in Logistics in Germany, the United States and the United Kingdom”. European Journal of Industrial Relations 28 (3): 363-382.
Hießl, Christina. 2022. “The Classification of Platform Workers in Case Law: A Cross-European Comparative Analysis”. Comparative Labor Law & Policy Journal 42 (2).
Holle, C.A. 1992. “Workers’ Participation and EC Legislation”. In Roger Blanpain, ed. Workers’ Participation: Influence on Management Decision-Making by Labour in the Private Sector. Deventer: Kluwer Law and Taxation Publishers, pp. 19-35.
Hooker, Josie and Lorenza Antonucci. 2022. “Improving the EU Platform Work Directive proposal. A contribution from emerging research findings”. OSE Paper Series, Opinion Paper No. 28. Brussels: European Social Observatory.
Howcroft, Debra and Birgitta Bergvall-Kareborn. 2019. “A Typology of Crowdwork Platforms”. Work, Employment and Society 33 (1): 21-38.
Howe, Jeff. 2008. Crowdsourcing – Why the Power of the Crowd is Driving the Future of Business. New York: Crown Publishing Group.
ILAW. 2022. Taken for a ride 2: Accelerating towards justice. Issue Brief.
ILO. 1997. Protection of Workers’ Personal Data. An ILO Code of Practice. Geneva: International Labour Office.
ILO. 1998. Declaration on Fundamental Principles and Rights at Work and its Follow-Up. Adopted at the 86th Session of the International Labour Conference (1998) and amended at the 110th Session (2022). Geneva: International Labour Office.
ILO. 2005. International Labour Conference, Report III (Part 1B), 93rd Session. Geneva: International Labour Office.
ILO. 2006. Employment Relationship Recommendation. No. 198.
ILO. 2016. Non-standard Employment around the World: Understanding Challenges, Shaping Prospects. Geneva: International Labour Office.
ILO. 2022. “Improving Workers’ Data Rights”. ILO Brief 1. Geneva: International Labour Office.
ILO. 2023a. World Employment and Social Outlook: Trends 2023. Geneva: International Labor Office.
ILO. 2023b. A normative gap analysis on decent work in the platform economy. GB.347/POL/1. 24 February. Geneva: International Labour Office.
Ivanova, Mirela, Joanna Bronowicka, Eva Kocher and Anne Degner. 2018. The app as a boss? Control and autonomy in Application-based management. Arbeit, Grenze, Fluss – Work in Progress 2. Europa-Universität Viadrina.
Katsabian, Tammy and Davidov, Guy. 2022. “Flexibility, Choice and Labour Law: The Challenge of On-Demand Platforms”. University of Toronto Law Journal forthcoming.
Kelly-Lyth, Aislinn and Adams-Prassl, Jeremias. 2021. “The EU’s Proposed Platform Work Directive: A Promising Step”, VerfBlog, 14 December. https://verfassungsblog.de/work-directive/, DOI: 10.17176/20211215-142502-0.
Koukiadaki, Aristea and Ioannis Katsaroumpas. 2017. Temporary contracts, precarious employment, employees' fundamental rights and EU employment law. PE596.823, Study for the Committee on Petitions. Brussels: European Parliament.
Kountouris, Nicola. 2012. “The legal determinants of precariousness in personal work relations: A European perspective”. Comparative Labor Law and Policy Journal 34 (1): 21.
Kountouris, Nicola. 2018. “The concept of ‘worker’ in European labour law: fragmentation, autonomy and scope”. Industrial Law Journal 47 (2): 192-225.
Krämer, Clara and Sandrine Cazes. 2022. Shaping the transition: Artificial intelligence and social dialogue. OECD Social, Employment and Migration Working Papers No. 279. OECD Publishing.
Kullman, Miriam. 2021. “Platformisation of work: an EU perspective on introducing a legal presumption”. European Labour Law Journal 13 (1): 66-80.
Kullman, Miriam, Ania Zbyszewska and Alysia Blackham. 2019. “Introduction”. In Miriam Kullman, Ania Zbyszewska and Alysia Blackham, eds. Theorising Labour Law in a Changing World: Towards Inclusive Labour Law. Oxford: Bloomsbury Publishing, pp. 1-16.
Laulom, Sylvaine. 2012. “Le droit européen des restructurations”. Le Droit Ouvrier 767: 420-426.
Lehdonvirta, Vili. 2016. “Algorithms that Divide and Unite: Delocalisation, Identity and Collective Action in ‘Microwork’”. In Flecker Jörg, ed. Space, Place and Global Digital Work, London: Palgrave Macmillan, pp. 53-80.
Mangan, David, Karol Muszyński and Valeria Pulignano. 2023. The platform discount: addressing unpaid work as a structural feature of labour platforms through labour law. Forthcoming, European Labour Law Journal.
Mason, Luke. 2020. “Locating Unity in the Fragmented Platform Economy: Labor Law and the Platform Economy in the United Kingdom”. Comparative Labor Law & Policy Journal 41 (2): 329-342.
McCann, Deirdre. 2005. “The role of work/family discourse in strengthening traditional working time laws: some lessons from the on-call work debate”. Law in Context 23 (1): 127-147.
McCann, Deirdre. 2016. “Travel time as working time: Tyco, the unitary model and the route to casualization”. Industrial Law Journal 45 (2): 244-250.
McCann, Deirdre. 2020. Temporal Casualisation and ‘Availability Time’: Mencap, Uber and the Framed Flexibility Model. DWR Research Paper 01/2020. Durham: Durham Law School.
Moyer-Lee, Jason and Nicola Kountouris. 2021. “The ‘Gig Economy’. Litigating the Cause of Labour”. In ILAW. Taken for a Ride: Litigating the Digital Platform Model. Issue Brief. March. Washington DC: ILAW.
Nagy, Csongor István. 2019. “Has the Time Come to Federalize Private Competition Law? The Autonomous Concept of Undertaking in the CJEU's Ruling in Case C-724/17 Vantaan kaupunki v. Skanska Industrial Solutions Oy, NCC Industry Oy, Asfaltmix Oy”. Maastricht Journal of European and Comparative Law 26 (5): 720-728.
Novitz, Tonia. 2020. “The Potential for International Regulation of Gig Economy Issues”. King’s Law Journal 31 (2): 275-286.
Piasna, Agnieszka. 2019a. Regulating uncertainty: variable work schedules and zero-hour work in EU employment policy. ETUI Policy Brief. Brussels: ETUI.
Piasna, Agnieszka. 2019b. “The Space for Regulation Beyond Borders? The Role of the EU in Regulating Zero Hours Work”. In Michelle O’Sullivan, Jonathan Lavelle, Juliet McMahon, Lorraine Ryan, Caroline Murphy, Thomas Turner and Patrick Gunnigle, eds. Zero Hours and On-call Work in Anglo-Saxon Countries, Singapore: Springer, pp. 179-194.
Piasna, Agnieszka. 2023. “Algorithms of time: how algorithmic management changes the temporalities of work and prospects for working time reduction”. Cambridge Journal of Economics forthcoming.
Piasna, Agnieszka. Wouter Zwysen and Jan Drahokoupil. 2022. “The platform economy in Europe – Results from the second ETUI Internet and Platform Work Survey (IPWS)”. Brussels: ETUI.
PPMI. 2021. Study to support the impact assessment of an EU initiative on improving working conditions in platform work. Luxembourg: Publications Office of the European Union, https://bit.ly/3wnr6aG.
Ponce Del Castillo, Aida and Diego Naranjo. 2022. Regulating Algorithmic Management: An Assessment of the EC’s Draft Directive on Improving Working Conditions in Platform Work. ETUI Policy Brief 8.2022. Brussels: ETUI.
Prassl, Jeremias. 2018. Humans as a service: The promise and perils of work in the gig economy Oxford: OUP.
Pulignano, Valeria, Agnieszka Piasna, Markieta Domecka, Karol Muszyński and Lander Vermeerbergen. 2022. Does it pay to work? Unpaid labour in the platform economy. ETUI Policy Brief 2021.15. Brussels: ETUI.
Rainone, Silvia. 2023. “Digital and Remote Work: Pushing EU Labour Law Beyond its limits”. In Nicola Countouris, Valerio De Stefano, Agnieszka Piasna and Silvia Rainone, eds. The Future of Remote Work, Brussels: ETUI.
Rani, Uma, Rishabh Kumar Dhir, Marianne Furrer, Nóra Gőbel, Aneliki Moraiti and Sean Cooney. 2021. World employment and social outlook: the role of digital labour platforms in transforming the world of work. Geneva: International Labour Office.
Rosenblatt, Alex. 2018. Uberland: How Algorithms are Rewriting the Rules of Work. Oakland, California: University of California Press.
Rubery, Jill and Agnieszka Piasna. 2017. "Labour market segmentation and deregulation of employment protection in the EU". In Agnieszka Piasna and Martin Myant, eds. Myths of Employment Deregulation: How It Neither Creates Jobs nor Reduces Labour Market Segmentation. Brussels: ETUI, pp. 43-56.
Stanciulescu, Alexandra. 2018. “The concept of undertaking in the European Union competition law”. Public Law Challenges of the Knowledge Society: 668-674.
Stanford, Jim. 2017. “The resurgence of gig work: historical and theoretical perspectives”. Economic and Labour Relations Review 28 (3): 382-401.
Stylogiannis, Charalampos. 2021. “Freedom of Association and Collective Bargaining in the Platform Economy: A Human Rights-Based Approach and an Ever-Increasing Mobilisation of Workers”. International Labour Review 162 (1): 123-145.
Supiot, Alain, M.E. Casas, P. Hanau and A. Johansson. 1998. “A European perspective on the transformation of work and the future of labor law”. Comparative Labor Law & Policy Journal 20 (4): 621.
Szpejna, Monika and Zahra Boudalaoui-Buresi. 2020. The scope of EU labour law. Publication for the Committee on Employment and Social Affairs, Policy Department for Economic, Scientific and Quality of Life Policies, European Parliament, Luxembourg.
Todolí-Signes, A (2021) “Making algorithms safe for workers: occupational risks associated with work managed by artificial intelligence”. Transfer: European Review of Labour and Research 27 (4): 433-452.
Tuominen, Mari. 2018. “Transparent and predictable working conditions – Initial Appraisal of a European Commission Impact Assessment”. Brussels: European Parliamentary Research Service.
Van Doorn, Niels. 2017. “Platform labor: on the gendered and racialized exploitation of low-income service work in the ‘on-demand’ economy”. Information, Communication & Society 20 (6): 898-914.
Veneziani, Bruno. 2019. “Article 27 – Worker’s Right to Information and Consultation within the Undertaking”. In Filip Dorssemont, Klaus Lörcher, Stefan Clauwaert and Mélanie Schmitt, eds. The Charter of Fundamental Rights of the European Union and the Employment Relation. Oxford: Hart Publishing, pp. 429–65.
Vieira, Tiago and Pedro Mendonça. 2023. “Beyond Misclassification: Algorithmic Management as Challenge to the Standards Employment Relationship”. Paper presented to ETUI conference on Future of Work: working with and through digital technology, 14 February, Brussels https://www.etui.org/sites/default/files/2023-02/Presentation_Tiago%20Vieira.pdf.
Weil, David. 2019. “Understanding the Present and Future of Work in the Fissured Workplace Context”. Russell Sage Foundation Journal of the Social Sciences 5 (5): 147-165.
Wood, Alex J. 2018. “Powerful times: flexible discipline and schedule gifts at work”. Work, Employment and Society 32 (6): 1061-1077.
Wood, Alex, J. 2021. Algorithmic Management Consequences for Work Organisation and Working Conditions JRC Working Papers Series WP No. 7.
Wood, Alex J., Mark Graham, Vili Lehdonvirta and Isis Hjorth. 2018. “Good Gig, Bad Gig: Autonomy and Algorithmic Control in the Gig Economy” Work, Employment and Society 33 (1): 56-75.
Wood, Alex J., Mark Graham, Vili Lehdonvirta and Isis Hjorth. 2019. “Networked but Commodified: The (Dis)Embeddedness of Digital Labour in the Gig Economy”. Sociology 53 (3): 931-950.
Wood, Alex and Vili Lehdonvirta. 2021. Platform Precarity: Surviving Algorithmic Insecurity in the Gig Economy, Working Paper for the “AI at Work: Automation, Algorithmic Management and Employment Law” online workshop, University of Sheffield, March 31.
Woodcock, Jamie. 2022. “Artificial intelligence at work: The problem of managerial control from call centers to transport platforms”. Frontiers in Artificial Intelligence 5.
Zahn, Rebecca. 2021. “Does stand-by time count as working time? The Court of Justice gives guidance in DJ v Radiotelevizija Slovenija and RJ v Stadt Offenbach am Main”. European Papers 6 (1): 121-124.
Acknowledgements
The authors are very grateful to Maria Helena André for her insightful comments on the draft version of this paper and Calvin Allen for his excellent language-editing.
