Reshaping the Legal Categories of Work

1 Introduction

While the policy debate on the regulation of digital platforms is picking up steam, the legal issues surrounding digital labor platforms are attracting increasing attention. For example, in November 2020, the German Federal Labor Ministry presented concrete proposals for “fair work in a strong platform economy.” 1

The term “digital labor platforms” 2 covers a wide range of platforms, from online clickwork platforms (e.g., Amazon Mechanical Turk) and bid-based design platforms for complex online crowdwork (e.g., 99designs) to platforms for on-site household services, transport, or food delivery (e.g., Helpling, Uber, or Glovo).3 The latter have particularly attracted labor lawyers’ attention. While the French, Italian, Spanish, and UK supreme courts have established that work on certain food delivery and transportation platforms is a form of “employment” or similar,4 the European Court of Justice 5 has been more hesitant, and a proposal to broaden the employee category in European law in Directive (EU) 2019/1152 also failed.6 Likewise, on November 3, 2020, Californian voters were asked to decide on Proposition 22, a legislative initiative backed by record amounts of money from Uber, Lyft, and Doordash.7 This proposition succeeded in undoing California Assembly Bill 5 (AB5) from September 2019, which had aimed to extend labor law coverage to workers on transport and delivery platforms.8

This article asks what this debate means for labor law. Is the category of “employment” that is central to labor law able to capture these new forms of work? Or is a fundamental paradigmatic change in the basic concepts underpinning labor law underway – a change requiring new categories?

The article addresses these questions by considering the entire spectrum of digital platform work. It starts by establishing the legal prerequisites to classifying digital platform work as employment (Section 1), draws lessons from organizational theory to explain the need for innovation (Section 2), and gives hints as to the possible consequences for the design of rights and obligations that could effectively protect workers on digital labor platforms (Section 3). It concludes by explaining how creating specific rules for digital labor platforms in their function as market organizers can lead to a more consistent and effective regulatory design (Section 4).9

2 Employment classification of digital platform work

One feature that most digital labor platforms share is that they classify platform workers as independent contractors rather than employees. Worker classification is important for digital labor platforms’ business models because the “employment” and “independent contracting” statuses entail widely differing legal and financial obligations. In most jurisdictions, these implications of classification extend not only to labor law but to other areas of law such as social security 10 or tax law.11 As far as labor law is concerned, in most jurisdictions, employee status means that the worker enjoys a full set of rights and the employer bears a full range of obligations. By contrast, classification as an independent contractor only entails the rights stipulated by the individual contract, i.e., what the platform has been able to dictate in line with its market position.

The strong interest of digital labor platforms in avoiding employment ­contracts is reflected in the terms and conditions they use in relation to their workers, which often include a variety of creative ways to frame them as ­independent contractors.12 While these instances of contractual “double-speak” have been convincingly dismantled,13 the scholarly debate around employment classification continues unabated. As many authors have lamented, it has been far from easy to reach conclusions in this regard.14

The following crossnational analysis will explain the central issues of these debates. This crossnational approach is possible because labor law is driven by similar demands and circumstances across the world – it seeks to regulate the relationship between worker and employer and is shaped by its economic circumstances. For this reason, the basic ideas underpinning labor law can be treated as functionally equivalent across jurisdictions.15 However, labor law is embedded in different legal and institutional contexts across national legal systems. For instance, any comparative study must account for the existence of further categories alongside the two primary categories of employment and independent contracting 16 (e.g., the Spanish “TRADE” category,17 the Italian “parasubordinazione” category,18 or the German “arbeitnehmerähnliche Person” category) 19. These variations form part of the backdrop for the discussion ahead, as the boundaries of the employment category differ depending on the relative range of any additional categories.

Notwithstanding this caveat, the following short summary of ideas will not concern itself with such differences and will instead focus on the basic common ideas. It will seek to explain the legal debate around digital platform work by describing how employment classification usually works (Section IA) and by showing how it has been applied in the context of digital platform work (Section IB).

2.1 The methodology of employment classification

Different jurisdictions describe the employment category in similar ways. The International Labour Organisation’s (ILO) Employment Relationship Recommendation No. 198 of 2006 implicitly explains the associated legal methodology and will serve as a valuable tool for this crossnational exercise, as it reflects the diversity of jurisdictions all over the world.20

The primacy of facts and the typological method of classification

Determining the existence of an employment relationship is a specific legal operation that differs from other legal operations. ILO Recommendation 198 shows this in Part II, which deals with the “Determination of the Existence of an Employment Relationship”. Here, Paragraph 9 establishes the principle of “primacy of facts” over contract, a rule that has also been dubbed the “economic perspective” or “business perspective.” 21 The principle of primacy of facts is designed to prevent “creative compliance” 22 with labor law obligations by lessening the leeway available to employers for avoiding or circumventing requirements under labor law. At its core, this approach invites us to look at the business model at stake rather than just read the contract.23

The primacy-of-facts principle has given rise to a specific methodology for classifying employment relationships. This “typological method” 24 consists of three steps. In the first step, certain criteria 25 describe or define the type (often referred to as a “category” in certain legal systems). In the second step, facts that constitute indicators of the type are identified. Finally, as a third step, the indicators are clustered 26 in an overall assessment.

Definitions of employment in classification tests

Using ILO Recommendation 198 as a generic guide, we can sum up the criteria for classifying work as “employment” as follows:

• Subordination, referring to “the fact that the work is carried out according to the instructions and under the control of another party” (Para. 12). This definition can be roughly identified with tests of “being bound by instructions” 27 or “control.” 28
• Integration “involves the integration of the worker in the organisation of the enterprise” (Para. 13a).
• Economic dependence on the employer. This criterion has to be treated with care in a comparative approach: Jurisdictions that do not rely on a “binary divide” 29 between employment and independent contracting often use economic dependence as the defining element of a third category.30
• Entrepreneurial opportunities, sometimes referred to as “business ­realities” or “economic realities,” concerning the worker’s opportunities to earn profit and the economic risks to which she is subjected.31
• Personal performance, finally, is addressed by the indicator that the work “must be carried out personally by the worker” (Para. 13a).32

In some jurisdictions such as Germany, the existence of an obligation to ­perform (Leistungspflicht) is an additional general contractual requirement.33

2.2 (How) Does digital platform work equal employment?

According to some of the jurisprudence and scholarly analyses on the classification of digital platform work in a variety of jurisdictions, workers on transportation platforms such as Uber and Lyft or on food-delivery platforms such as Deliveroo and Glovo have already been classified as employees or workers, or, at least, as economically dependent persons.34

Looking at the tests, indicators, and the ways they are clustered when classifying these platform workers,35 we find that the obvious empirical situation – workers are not autonomous – does not easily lend itself to translation into labor law. The French Cour de Cassation in its November 2018 (Take Eat Easy) and March 2020 (Uber) decisions ultimately had to use new indicators 36 for assessing platform work as employment.37 The same is true for the UK Supreme Court’s judgment of February 2021.38 The Spanish Tribunal Supremo, in its September 2020 ruling (Glovo), explicitly acknowledged the need to adapt criteria to a “new reality.” 39 The Italian Corte di Cassazione had already reached a similar conclusion in January 2020 with regard to the food delivery service Foodora.40 At the same time, and in contrast to these openings, the European Court of Justice, in April 2020, ruled on the Yodel delivery service by a reasoned order 41 instead of a formal judgment; this indicated that it did not consider it useful to think about modifying criteria. Unsurprisingly, this is the decision that did not acknowledge the existence of “employment.” 42

In contrast to transportation and food delivery, crowdwork has hardly ever been deemed employment. With a lack of court cases on the matter, scholars have taken up the issue and usually classified it as independent contracting,43 or – where this category exists – as industrial homework/piecework (German “Heimarbeit”) 44 The German Federal Labor Court, in December 2020, seems to have been the first supreme court to rule on crowdworking, and rather surprisingly established status of the plaintiff (a platform worker) as an employee. Here again, an innovative reinterpretation of criteria and indicators for employment was needed in order to achieve this result; rather than looking at “specific instructions,” the court focused on the incentives created by the rating system and by the way the platform presents the tasks to the worker.45

These decisions are evidence of the difficulties that must be worked around by those wishing to classify digital platform work as employment. First, when a platform construes the legal situation as one of contracting between a customer and worker, this contractual situation must be disregarded on the grounds that it is “artificial and unproductive” 46 in order to classify the arrangement as one of employment. Second, traditional indicators usually focus on substantive questions of subordination, economic dependence, and organizational integration. For this reason, court decisions that have classified digital platform work as employment have had to rely on new and mostly innovative indicators. To date, four characteristics of digital platform work have played a pivotal role here:47

• Obligatory use of platform apps, specific soft- and hardware, or other related tools (i.e., “app-based management”48);
• Rating and feedback mechanisms;49
• Qualification requirements and the assignment of tasks to particular workers;
• The economic positions of platforms and workers (i.e., access to markets, entrepreneurial opportunities).

These indicators refer to forms of control that have in the past not been widely used in employment classification. Instead of looking at direct control, subordination, and organizational mechanisms of direct access to workers’ bodies, they focus on the “indirect control” exerted via mechanisms that structure action and create motivation and commitment.50

However, although some courts have succeeded in reinterpreting criteria (at least in cases of transport and delivery work), these conceptual challenges have not yet been sufficiently reflected in labor law theory.

3 Deconstructing and reconstructing labor law

This is not at all a new problem, and the gig economy may only be an example of the “fragmentation” of organizations,51 the “fissuring” of firms,52 “post-bureaucratic” forms of organization,53 “fluid organisations,” 54 or the new “digital mode of production”.55 Consequently, there is an intensive debate on how to deconstruct and reconceptualize the categories of labor law in order to capture indirect forms of management and control. The following section will explain the basic conceptual approaches.

3.1 Theoretical approaches to labor law classification

Embedding labor law in human rights approaches

The Supiot Report of 2001 is an influential example of an attempt at deconstructing and reconceptualizing the purposes and general normative ideas of labor law. It was the work of a comparative research group of European labor lawyers lead by Alain Supiot 56 that set out to answer, inter alia, the European Commission’s question: “Is there a need for a ‘floor of rights’ dealing with the working conditions of all workers regardless of the form of their work contract?” 57

The Supiot Report proposed regulating work (including unpaid work) in four circles: dependent employment in the narrow sense, professional work, non-professional (unpaid) work, and (in the outermost circle) any work activity.58 Mark Freedland and Nicola Kountouris, in their own reconstruction suggested a mobile system of elements. In order to develop “neutral” or “baggage-free” analytical concepts, they separated two analytical elements, namely the “personal work relation” and “personal work nexus,” from contractual analysis,59 equating their “personal work relation” to the Supiot report’s “statut professionel” 60

These comprehensive concepts are ultimately capable of identifying commonalities shared by any kind of work, i.e., justify universal rights at work. Consequently, they define a basic element, which they link with a human rights approach.61 Embedding labor in universal human rights 62 is also an objective identified in the claim “labour is not a commodity” or, rather, “labour is a fictive commodity.” 63 By emphasizing that work is inseparable from the person performing it,64 the phrase focuses attention on human dignity while at the same time acknowledging the empirical fact that labour law “underpins the creation of labour power as a commodity, and regulates the resulting social and economic relations.” 65 This is why Amartya Sen’s notion of “capabilities” has been so attractive 66 for this endeavor: It “provides a framework for debating which labour and social rights ought to be considered fundamental”.67

However, some doubts remain as to what reconstructing labor law by way of human rights can achieve in our context.68 The specificity of labor law lies exactly in its ability to capture economic realities, power, and market failures and to promote social solidarity.69 Therefore, the necessary reconnecting of specific rights and obligations (wage, working time, collective organization, etc.) goes beyond identifying universal rights. In order to link the legal character to legal consequences,70 we will have to look at the specific rationales of labor law.

Power and dependence as the defining rationale for labor law rights and obligations

My own proposal for a reconceptualization of labor law categories builds on an acknowledgment of the function of labor law in controlling “private power.” It starts by identifying the different socio-economic sources of power at work – sources and phenomena of power that may, could, and should give rise to specific regulations.71 These are not limited to market power, but include, as in any long-term relationship, economic lock-in problems of specific investments that create barriers to exit in cases of conflict.72 Most importantly, however, work in organizations possesses a unique feature that creates specific power for the employer. Hugh Collins, in his 1986 article on the contract of employment, aptly termed this problem as one of the difference between “market power” and “bureaucratic power”: “even with reduced inequality of bargaining power, the social dimension of subordination remains.” This managerial power is the result of how work is organized: “An employee normally joins a bureaucratic organisation and will be allocated a particular role, defined by the rules of the institution.” 73

Rolf Wank has advanced a similar idea, first in his 1988 postdoctoral thesis and in numerous articles since.74 He also starts from the assumption that legal consequences should have to be justified in view of specific problems and states that it is managerial control exerted by an employer over a worker that justifies most of the protection against control and unfair risks afforded by labor law. His basic contribution to labor law theory, however, is his look at market alternatives to legal rights: Wank expressly defines the dependence labor law should react to as the situation of not being able to care for oneself economically because one is bound to work for the benefit of others. We find a similar idea in the second aspect of Davidov’s “purposive approach” to labor law: He defines the vulnerabilities at the heart of labor law as subordination (in the sense of democratic deficits) and dependence (in the sense of the inability to spread risks).75

3.2 Digital platforms: hierarchy, networks, and market organization

In view of these debates, an approach that entails looking at the ways in which digital labor platforms create dependence, control, and the organizational power that gives rise to the need for specific labor law rights and obligations seems especially promising. Organization theory can help to define this further.

Organization theories

Organization theory aligns with labor law in its attempt to explain how the coordination of goods, services, and work by organizations differs from the coordination by market mechanisms. In other words, organization theory explains and draws a line between “making” a product or service (with employees) versus “buying” it on a market (from an independent contractor).76 It is thus not a coincidence that Ronald Coase, who developed the theory of the firm for organization theory, used the (then-common) legal criteria for worker classification. He described the “firm in the real world” with “the master’s right to control the servant’s work, […] of being entitled to tell the servant when to work (within the hours of service) and when not to work, and what work to do and how to do it […].” 77 The “binary divide” 78 between employment and independent ­contracting is obviously not something rooted in contract law but is rather a result of an economic and institutional duality.

Since Coase, organization theory has developed significantly. It now describes the coordinating mechanisms used by organizations in more detail, using elements of membership, hierarchy and control, rules, monitoring, and sanctions.79 The concept of hierarchical organization has, however, also been questioned in organization theory and been criticized for only reflecting “the social reality of the employment relation in advanced industrialised societies.” 80

Digital labor platforms as organizations

There have been various attempts in organization theory to make sense of the developments of “post-bureaucratic” forms of organization summarized above.81 In relation to digital platforms and the gig economy, the concept of “network” 82 (which replaced the term “clan” 83) has been particularly influential. It is intended to describe any nonmarket and nonorganisational form of coordinated “inter-organisational value-creation.” 84 As such, the network concept is well suited to describing and analyzing the sharing economy and horizontal or cooperative systems.85 It is less useful, however, when it comes to identifying power structures.

In relation to digital labor platforms, Stefan Kirchner and Elke Schüßler have developed a more promising approach.86 Kirchner and Schüßler use the term market organizer to describe the specific functions of digital labor platforms for organization theory,87 in line with the idea of “partial organisations.” 88 First, these platforms open up markets that “may lower the costs and risks of setting up as an ‘independent expert’ and make it easier to access work.” 89 Second, they fulfill specific functions of trust-building and quality control within these markets. This is why their feedback and reputation mechanisms are such important features,90 and this is why they ultimately do need some degree of control over market access 91 and control over their workers.92

Delimitation: Conflicting concepts of “market organizers”

If characterized as market organizers, digital labor platforms should be analyzed according to functions like managing transactions, establishing a system of reputational feedback and indirect control, providing additional information to users (such as security or personal identity checks), and processing money exchange, thereby rendering it impossible for their workers to gain independent market access.93 Understood as “gatekeepers”94 and “private legislators,”95 digital platforms would embed the rules for how to participate and act in the competition in their design and architecture.

If “market organization” served as the basis for a new labor law category, these characteristics of indirect management could be used as indicators.96 This would go well beyond what several Republican-governed US states accomplished in 2018 when they passed laws introducing the category of “market contractor”, retaining a binary model of employment classification by defining the new category as a nonemployee.97 Contrary to this approach, introducing the idea of the market organizer would not simply address the formal structure of market mediators but would instead identify those features that form the basis of the specific power imbalances these market organizers create.

4 Dangers and benefits of specific regulation

This theoretical framework could be used for different purposes; in many instances, it could be used by courts or legislators98 to more systematically advance concepts of employment in the strict sense – if judges continued to innovatively develop this category.99 This could be useful for transport, delivery, or household services, i.e., for cases of digital platform work that exhibit more than just one feature and indicator of control.

The framework could also serve as the basis for specific definitions in particular sectors 100 – or even for general rules on digital platform work, as is evident in the German Labor Ministry’s approach.101 For these purposes, it makes sense to introduce a new category as starting point for specific regulation.102 However, such strategies have been heavily disputed.

4.1 Cons: the dangers of deregulation

The main argument against new and intermediate categories is that digital platform work is just a modern form of precarious atypical employment. Based on this assumption, some conclude that the introduction of further forms of contracts and protection is nothing but deregulation.103 This is an argument that not only favors protecting platform workers but is also concerned with a “level playing field” with respect to both employment and tax law as well as consumer protection or social security contributions.104 After all, crowdworking has often been advertised as an alternative to employment 105 – this renders the assumption that the gig economy could be used as an instrument to replace standard employment all the more credible.

Nevertheless, experiences in states that have additional employment categories do at least not give grounds to clearly advise against the strategy proposed here.106 More importantly, the most persuasive argument against persevering with the existing binary is its ineffectiveness against deregulation. Opportunities for evasion already exist 107 – this is partly in spite of legal regulation and due to the organizational choices opened up by technological developments. The rift between employment law and general contract law tends to drive dynamics of circumvention.108 This is the idea behind Simon Deakin’s statement that the “worker concept preserves the contract of employment only at the expense of diminishing the scope of application of the core model.” 109

4.2 Pro: the benefits of fit-for-purpose rules

Equating gig-economy work and traditional employment may be right and wrong at the same time. On the one hand, work on digital platforms does often entail control, a lack of autonomy, and levels of precarity comparable to low-standard employment. On the other hand, from a legal perspective, it almost as often differs from standard employment in some important respects.

A closer look at the internal dynamics of employment classification suggests that the very insistence on a concept that may not really fit the problem could be backfiring. After all, the effectiveness of labor law classification is based on the methodological principle of the primacy of facts which, in turn, is based on the assumption that there are certain features of work coordination and organization that a firm cannot easily change. It works “by reference to social and economic criteria which reduce as far as possible the influence of the employer’s choice of form.” 110 Consequently, the criteria and indicators used to determine employment classification should accurately reflect digital labor platforms’ actual organizational mechanisms. Effective regulation of digital platform work should be designed to address platforms’ specific business models and organizational principles as directly as possible.

5 Labor law regulation for market organizers

If lawmakers wish to link categories to legal consequences,111 they should arguably define market organizers with reference to organization theory, as proposed here, as this can help identify which specific rights and obligations would be adequate. This task goes far beyond what can be done in this article. The following ideas therefore focus on issues of governance and the forms of regulation that could address the specific structures of market organizers. After explaining some of the general ideas on the governance and regulation of digital platforms (5.1), including techniques of organizing accountability (5.2), I highlight issues of collective action as a central concern for labor law (5.3).

5.1 Digital platforms as objects of legal regulation

Digital platforms have come to be seen as the characteristic organizational form of “informational” or “digital” capitalism.112 The questions of whether and how to regulate these platforms has therefore become one of the primary concerns in legal debates on the digital economy. These concerns have mostly arisen due to the realization that digital platforms are more than just technological instruments or “matchmakers.”113

Regulators’ concerns in relation to digital platforms in general are manifold: From a consumer law background, the EU’s “P2B-regulation” 114 addresses the business environment for smaller businesses and traders on online platforms. For competition (antitrust) law, EU lawyers have recently suggested creating stronger antitrust control of online platforms, given their extreme returns to scale, network externalities and the crucial role of data for developing new services and products.115 Similar discussions are taking place in the United States.116 There is also an important human rights discourse: The European Council’s 2018 policy recommendations for regulating digital platforms for example, represents a comprehensive attempt to outline digital rights and establish the roles and responsibilities of digital platforms.117

Combined with the debates in data protection and property rights law, these have contributed to an emerging digital law discourse.118 Access to data, facilitation of switching and multi-homing, data portability, and clear rules for ratings are the basic pillars of the digital law framework, along with privacy, data protection, and access to effective remedies.119

As a subgroup of digital platforms, digital labor platforms are implicitly implicated in these digital law discourses. The P2B Regulation (EU) 2019/1150, although not designed for these platforms,120 does offer rules beyond those concerned with transparency that could be applicable to digital labor platforms, such as restrictions with regard to termination of contract or downgrading in rankings.121 Provisions on complaint management, dispute resolution, and the right of associations to take legal action would also cover trade unions. Overall, however, rules on the transparency of ranking criteria, on access to data, and on the portability of reputational data are rather designed to enable a certain minimum level of entrepreneurial action 122 rather than
protecting workers from the inevitable organizational power of platforms.

5.2 Organizing accountability through procedural and reflexive law

Digital law has learned some lessons on a second issue, namely accountability. Since its emergence, debates on regulating digital platforms have been concerned with organizing accountability and legal responsibility for digital platforms. The triangular character of digital labor platforms has likewise been an important issue for labor law. However, specific labor law instruments on triangular relationships such as regulations on temporary work 123 or employment agencies 124 have tended to lead to impasses, because such regulations concern themselves with equal treatment with respect to the hirer and lead the focus away from the agency.125 They can therefore be at best of marginal interest when it comes to holding digital labor platforms directly accountable.

When it comes to regulating digital labor platforms, looking at specific functions of market organizers could add something to the debate. First, the distinction between “active” and “passive”/“neutral” platforms126 has been instrumental in ascertaining levels of activity in coordinating work and providing services on digital platforms and using it to allocate legal accountability.127 Second, specific concepts for sharing responsibilities 128 that propose procedures akin to due diligence procedures have been developed.129 With regard to platforms that are not hierarchically structured, it may be a good idea to use procedural and reflexive rules to organize accountability for substantial rights, such as the rights to health and safety according to activity, as this would make it possible to hold platforms accountable for organizing compliance with the law.130

5.3 Collective action and antitrust law

Two key issues arise as central concerns when focusing on organizing accountability: participation and workers’ collective representation.131 Given the lack of specific and effective regulation or litigation to date, collective action by workers is, both theoretically and practically, a powerful instrument for counter-balancing digital platforms’ organizing power.

Notwithstanding international law that includes self-employed solo-entrepreneurs in the scope of collective bargaining,132 collective bargaining practiced by “independent contractors” runs up against obstacles in antitrust law, where the growing number of collective agreements 133 and price regulation 134 for platform services will potentially be deemed illegal.135 European Union Competition Commissioner Margrethe Vestager has already acknowledged the necessity of “[making] sure that there is nothing in the competition rules to stop those platform workers from forming a union,” 136 and the European Commission’s consultation on the Digital Services Act Package has identified collective organization of workers as a central challenge on digital labor platforms.137

Lessons from labor law can contribute to identifying the issue: Antitrust law has been based on a binary model of organization versus market that is similar to the one that has proven so complicated in labor law. It draws on similar assumptions to those underpinning traditional labor law classification when it bans rule-setting on “free markets” and allows rule-setting within coordinated organizations (“firm immunity”). Consequently, independent contractors cannot benefit from neither the labor exemptions nor the firm immunities currently provided for in competition law.138

Yet, labor lawyers could learn from recent discussions in competition/antitrust law when seeking to identify criteria and indicators that distinguish digital labor platforms and other market organizers from hierarchical organizations.139 As for EU law, the European Court of Justice’s decision in the FNV Kunsten case is an example of such an endeavor: While seemingly sticking to a binary system of employment classification, the ECJ in this case implicitly used criteria developed in competition law like the category of the “independent economic operator” to identify the scope of collective bargaining. 140 The assumption that there are no intermediate categories in European competition law 141 is thus outdated and disproven.142 Economic actors like self-employed workers who depend on market organizers’ control of market access, contractual terms and conditions, and digital working tools should enjoy rights to collective action and immunity from competition rules.143

6 Summary and Conclusion

All of these issues require further and more detailed analysis for each jurisdiction. Comprehensive reviews such as the Hamilton project in the United States,144 the UK’s Taylor Review,145 or the German Labor Ministry’s White Book “Arbeiten 4.0” 146 have all proven how much classification of work relationships implicates specific political, economic, and social institutions and contexts. Regardless of jurisdiction, however, the regulation of digital labor platforms must be thought of as part of a broader debate on law and political economy, on a labor constitution for the digital age and informational capitalism,147 and as a way of re-institutionalizing work by directly tackling the diversity of today’s work relationships.148

In this vein, the framework proposed here could represent a way out of the impasse that the debate around classification seems to have reached: Even authors who strongly advocate for classifying digital platform work as employment rarely support a comprehensive application of employment law in the strict sense. Consequently, for the most part, relevant policy debates have treated the question of how to classify workers as employees and how to allocate specific rights and obligations for digital platform work as two separate issues.149 However, a comprehensive conceptual approach would have to address the organizational character of digital labor platforms as market organizers on two levels at once: First, it would have to modify the criteria and indicators for classifying workers, either in the employment category or in a new category, to capture indirect mechanisms of worker control, such as feedback and rating systems. Second, the rights and obligations associated with labor law, as well as the participation and governance structures, would have to be reformulated to address indirect control and the social dynamics of virtual workplaces. It is time to acknowledge the paradigmatic changes for labor law that are underway in the working world.

Funding and acknowledgements

The research reported in this article received funding from the Hans Böckler Foundation (grant 2017-464-2) and Fritz Thyssen Foundation (grant 10.15.1.003SO). Thanks go to Jochen Koch, Anna Schwarz, Joanna Bronowicka, Mirela Ivanova, Sean King, Isabell Hensel and Daniel Schönefeld for these cooperations. I also thank the anonymous reviewers and the language editor of the Weizenbaum Journal of the Digital Society as well as Allison West who did valuable language editing on a former version of the article.

Date received: April 2021

Date accepted: May 2021