A dispatch is a brief essay, ordinarily not to exceed five printed pages, describing a significant development in national labor law: legislative, judicial, administrative. The importance of the development in domestic context should be explained; the reasons for transnational interest might be suggested.

Proposed dispatches should be submitted to the Journal’s editorial office at law-cllpj@illinois.edu and, if approved, will be posted on the Journal’s website. The titles and authors of Dispatches appearing from one number to the next will be published in the Journal.


Dispatch No. 47 - Germany - "The German Lieferkettensorgfaltspflichtengesetz (LkSG) – a step towards more humanity in supply chains?"

by Helene Langbein

June 2023


Dispatch No. 46 - Germany - "Social Sustainability in Global Value Chains—New Supply Chain Act in Germany"

by Thomas Klein

March 2023


Dispatch No. 45 - Chile - "Chile's Legislative Solution For Platform Work: Does it Fit the Bill"

by Rodrigo Azócar Simonet, Jorge Leyton García

October 2022


Dispatch No. 44 - Brazil - "Employees Hired in Brazil and Working from Home for the Whole World"

by Murilo Meneghetti Nassif

September 2022


Dispatch No. 43 - New Zealand - "Pay Equity Bargaining in New Zealand"

by Amanda Reilly, Avalon Kent, Annick Masselot

August 2022


Dispatch No. 42 - Turkey - "What Has Been Done after the Soma Disaster and is it Adequate to Create Safe Working Conditions in Turkey?"

by Merve Kutlu Mutluer

April 2022


Dispatch No. 41 - Spain - "The Spanish Labor Reform of 2021: A Model to Follow Due to Its Origin and Objectives"

by Lourdes Mella Méndez

March 2022


Dispatch No. 40 - European Union - "The EU Proposal for a Directive on Improving Working Conditions in Platform Work"

by Jelena Starcevic

January 2022


Dispatch No. 39 - European Union - "The AI Act and Algorithmic Management"

by Aislinn Kelly-Lyth

October 2021


Dispatch No. 38 - Norway - "The 'Back and Forth' in the Protection of (Collective) Labor Rights Under the ECHR Continues: The Holship Case"

by Charalampos Stylogiannis

September 2021


Dispatch No. 37 - Argentina - "How Did a Food-Delivery Platform's Judgement Transform Freedom of Association Into a Second-Class Right?"

by Kevin Hartmann-Cortés

August 2021


Dispatch No. 36 - Spain - "The Legal Framework of Platform Work in Spain: The New Spanish 'Riders' Law'"

by Daniel Pérez del Prado

July 2021


Dispatch No. 35 - Italy - "Demystifying Flexibility, Exposing the Algorithmic Boss: A Note on the First Italian Case Classifying a (Food-Delivery) Platform Worker as an Employee"

by Antonio Aloisi

June 2021


Dispatch No. 34 - Belgium - "Cheap and Flexible Labor for Non-Profits: How the Belgian ‘Act on Auxiliary Income’ Evolved Into an ‘Act on Association Work’

by Mathias Wouters

May 2021


Dispatch No. 33 - Russia - "The Reform of Remote Work in Russia"

by Olga Chesalina

April 2021


Dispatch No. 32 - Argentina - "The Right to Fair and Satisfactory Working Conditions Through the Lens of the Inter-American Court of Human Rights"

by Renan Kalil & Mauro Pucheta

March 2021


Dispatch No. 31 - United States - "Proposition 22: A Vote on Gig Worker Status in California"

by Miriam A. Cherry

February 2021


Dispatch No. 30 - Spain - "Notes on the Spanish Supreme Court Ruling that Considers Riders to be Employees"

by Adrián Todolí Signes

December 2020


Dispatch No. 29 - Brazil - "Coronavirus and Turbulence in Brazilian Labor Law: An Analysis on the Main Legal Changes During the Pandemic"

by Catharina Lopes Scodro & Olívia de Quintana Figueiredo Pasqualeto

November 2020


Dispatch No. 28 - Austria - "The Problem with Highly Mobile Workers in the EU: A Critical Analysis of the Court of Justice of the European Union Judgment in Dobersberger Regarding the Posting of Workers"

by Julia Heindl & Magdalena Lenglinger

October 2020


Dispatch No. 27 - Poland - "Poland on the Path towards Whistleblower Protection"

by Łucja Kobroń-Gąsiorowska

September 2020


Dispatch No. 26 - United States- "The Landmark Bostock Decision: Sexual Orientation and Gender Identity Bias in Employment Constitute Sex Discrimination Under Federal Law"

by Susan Bisom-Rapp

August 2020


Dispatch No. 25 - Spain - "A New Judgment from the Court of Justice of the European Union on Temporary Hiring in Spanish Public Employment, Based on Abuse of Law"

by Edurne Terradillos Ormaetxea

July 2020



Dispatch No. 24 - United Kingdom - "The Court of Justice of the European Union’s Order on B v Yodel Delivery Network"

by Ricardo Buendia*

June 2020


I.     Facts of the Case

On 22 April 2020, the Eighth Chamber of the Court of Justice of the European Union (CJEU, or the Court) made a reasoned order in response to a request for a preliminary ruling. The request came from an employment tribunal in Watford (the ET), United Kingdom (UK) in the proceedings brought by “B,” formally a self-employed independent contractor, against ‘Yodel Delivery Network Ltd’ (Yodel), a parcel delivery company. This case is about B’s professional status in his employment relationship with Yodel.

B is a courier working who has worked exclusively for Yodel since 2017 under what was descried as a “courier services agreement.” This contract, among other things, stipulates:

A.   B is deemed a self-employed independent contractor of Yodel.

B.    B can use substitutes or subcontract others for the performance of the services provided they fulfil qualification and abilities at least equivalent to what Yodel requires of their couriers. On this point, Yodel reserves a power of veto. In any case, B remains personally liable for any act of the substitutes or subcontractors.

C.    B can work for third parties, including direct competitors of Yodel.

D.   Yodel may choose not to use the services of B and B is not obliged to accept any parcel for delivery. Also, B can fix unilaterally a maximum number of parcels to be delivered.

E.    B gets paid a fixed rate per parcel delivered, and the price is set according to the place of delivery.

In addition, it is stated in the order that:

A.   B has received training from Yodel to be able to perform part of his duties.

B.    Couriers use their own vehicle and phone to deliver parcels handled by Yodel.

C.    B can choose the number, the route to take, and the schedule to deliver the parcels. The exceptions to these conditions are fixed-time parcel delivery and a general timeframe to delivery parcels, between 7.30 and 21.00.

B brought a claim against Yodel in the ET in Watford. B argued that he ought to be deemed a worker and not as self-employed for the purpose of EU Directive 2003/88/EC of the European Parliament and of the Council of 4 November 2003 concerning certain aspects of the organisation of working time (Working Time Directive, or WTD).

II.     Preliminary Ruling and the Reasoned Order

To address this issue, the Watford ET sought a preliminary ruling from the CJEU. This procedure is regulated under Article 267 of the Treaty on the Functioning of the European Union (TFEU). This Article aims to ensure the uniform interpretation of EU law by all Member States.[1] It does so by giving preliminary rulings on the interpretation of treaties, or the validity and interpretation of acts of the institutions, bodies, offices or agencies of the Union.[2]

The CJEU then examined the issues brought by the ET. First, the Court correctly declined to opine on the facts of the case as this must lie in the hands of the national court.[3] As will be seen, that feature is relevant in the case at hand. Second, the Court correctly engages with issues around the interpretation of the WTD, which is considered part of the EU legal order, as are the EU treaties. A preliminary ruling of the Court on the correct interpretation of the WTD in a case like this is relevant as it has the potential to shed some light on the determination of the meaning of “worker” more generally where it figures in EU law, and specifically in relation to non-standard employment status in the EU. It is therefore a profoundly controversial issue.

In this case, the Court decided to make a reasoned order. Orders are a faculty for the CJEU based on Article 99 of the Court’s Rules of Procedure[4] to rule by “referring to a previous decision or earlier case law where a question referred is identical to one that has already been answered, or where the answer to the question can be clearly deduced from prior case law.”[5]  The Court deemed the question brought by the Watford ET to be answered through an order on the basis that the question can clearly be deduced from existing case-law or that the question does not admit a reasonable doubt.[6]

III.     The Question and the CJEU Order

In essence, the Watford employment court asked the CJEU whether the WTD is to be interpreted as applicable to a “worker” when the person concerned:

A.   Does not have to perform their services personally, and in particular:

i.         can subcontract or use substitutes to perform the agreed services.

ii.         has, in fact, contracted subcontractors or substitutes.

B.    Has the capacity to reject tasks given by the employer and the putative employer is not obliged to offer jobs to the worker.

C.    Can provide services to third parties during the contract period, including the provision of services to direct competitors of the putative employer or has, in fact, worked for third parties.

D.   Can fix their schedule under a given timeframe.

This question raises issues as to whether the applicability of the WTD is a good fit with UK legislative requirements that determine access to employment rights. In the case at hand, what matters was whether B could be deemed a worker to apply the WTD.

There is a difference between the UK and the EU regarding employment classification. On the one hand, according to section 230(3) of the 1996 UK Employment Rights Act (ERA), “worker” [or limb (b) worker] is a sub-category of self-employed who carries their work out as part of someone else’s business, but who is not considered an employee. In this regard, limb (b) workers enjoy some rights of employees but are deemed self-employed in other respects, for example regarding protection from dismissal.[7] On the other hand, the EU has developed a concept of worker that is based on the binary divide between worker – a person in an employment relationship – and the self-employed or independent contractors.[8] In this case, the concept of “worker” is more related to the UK idea of an employee.

To answer the question of the Watford ET , the CJEU reminded the ET that the WTD does not have a specific concept of worker.[9] Nevertheless, there is an autonomous concept of worker developed through the CJEU case-law and applicable in EU law.[10] The EU concept of worker would not be applicable at the national level when considering rights that are not governed by EU law. On the other hand, an EU concept of worker would be relevant when there is a set of EU norms at stake, such as the WTD.

Also, the Court recalls that, even where national legislation classifies a person as self-employed, this person can be reclassified as a worker in accordance with EU law if their independence is merely theoretical, hiding a real employment relationship.[11] This development is interesting as the CJEU expands the EU definition of worker and departs from cooperation to impose somewhat hierarchical coordination criteria to other national courts.[12]

Further, the CJEU has established that an independent service provider is associated with more flexibility in choosing the type of work, the schedule, the methods and the tasks to be performed, the place to perform the services and the freedom to hire own collaborators.[13] The more the putative employment relationship fulfils this test, the more independent the worker will be.

Beyond the abstract analysis of the EU concept of worker by the CJEU, the Court recalls that there is a need to analyse whether the independence of the worker “does not appear to be fictitious and it is not possible to establish a relationship of subordination between that person (the worker) and the putative employer.”[14] In this regard, the Court emphasises the need to evaluate the specific circumstances of the case to assess the nature of the employment relationship. For the CJEU, the EU definition of worker must be considered by the remitting court when interpreting the WTD.[15] Further, to assess the categorisation of B as a worker, the remitting court must “base that classification on objective criteria and make an overall assessment of all the circumstances of the case brought before it.”[16]

Despite the CJEU reinforcing the need to evaluate the reality of the employment relationship, the Court then carried on an analysis of the elements that would define worker under EU law, based on the fact that B appeared to have a “great deal of latitude in relation to his putative employer.” Based on the elements provided by the referring ET, the Court concludes there are not enough elements in the relationship between B and Yodel to be considered a relationship of employment. Therefore, the service agreement between the parties was not fictitious as B seems to enjoy great freedom against the putative employer. The fact that B could hire substitutes or subcontractors was a possibility reserved, in principle, for service providers and not subordinate workers. Also, the capacity of B to reject assigned tasks by Yodel, to establish a limit for the tasks he is willing to execute, the fact that he could provide services to other companies -including directly competing companies-, and the establishment of his own schedule under certain limits, are indicia to arrive to this conclusion. A critical analysis of this outcome will be carried out in the next section.

IV.     Some Considerations on the Decision

First, the CJEU reaffirms that the determination of an employment relationship depends, ultimately, on the analysis of each case. Also, it can be argued that the CJEU identifies dynamic elements to determine an employment relationship that must be read in context. For instance, the Court implicitly reminds national courts to calibrate the meaning and level of freedom in a specific working relationship, thereby evaluating the existence of a hierarchy,[17] “the nature of the activities concerned and the relationship of the parties involved” to identify an employment relationship.[18]

This approach is in line with ILO Recommendation No. 198.[19] According to this Recommendation, the determination of an employment relationship “should be guided primarily by the facts relating to the performance of work… notwithstanding how the relationship is characterized in any contrary arrangement, contractual or otherwise, that may have been agreed between the parties.”[20] Also, the CJEU view is aligned with the doctrine of the primacy of reality, which aims to give preeminence to the reality of working conditions over formal arrangements between the parties or the classification under domestic law.[21] This path of the Court seems appropriate to the case at hand.

Second, the business model of Yodel is a neat fit with the so-called “gig-economy.” In the gig-economy, to maintain a large pool of workers ready for the next gig without spending additional resources on them, companies usually consider their workers self-employed.[22] Yodel relies on a large pool of couriers who are deemed independent self-employed service providers. In this view, as self-employed, B could reject tasks assigned by Yodel without affecting the operation of the business, because it is likely that another person of this large pool of workers will complete the offered task

It is surprising that the Court discusses subordination based on the factors it does, especially considering the strategies of some platform companies to mask an employment relationship behind self-employment and entrepreneurship.[23] This view of subordination developed by the Court creates some dissonance between the abstract interpretation given by the CJEU on some elements of this judgment to identify subordination,[24] and therefore an employment relationship, and the objectives of the Directive on Transparent and predictable working conditions (TWCD).[25]

The TWCD sets new rights for all workers, especially those who work in more precarious jobs.[26] Ostensibly, the TCWD defers to the established case law of the CJEU on worker status, but certain provisions of the Directive are highly relevant for the purposes of the facts of the Yodel Delivery case. In this regard, articles 10.1 and 10.2 of this Directive state, given certain conditions, when working patterns are entirely or mostly unpredictable, the worker shall not be required to work. Moreover, the worker has the right to reject tasks without adverse consequences. When comparing TWCD with the case at hand, it can be said that the capacity of B to reject tasks given by Yodel should not be indicia suggestive that there is no subordination because TWCD regulates this issue for workers when schedules are unpredictable. It seems that the ability to reject work is protected under TWCD in the context of a working relationship.

Also, the capacity of B to provide services for other companies does not entail that there is no subordination of B to Yodel. This conclusion emanates from article 9 TWCD as it guarantees to workers the possibility to work for others under certain conditions. Also, the capacity of a worker to provide services to other companies can be conceived in more standard employment relationships, and that does not seem prima facie to alter the legal status of the parties.

In addition, the possibility to hire subcontractors or substitutes by B should not be indicia to discard subordination. There is evidence from national courts, that the fact that Yodel reserves a right to reject subcontractors or substitutes based on an alleged level of competence has not just the potential to become arbitrary, but also to make this faculty of B merely theoretical.[27] It seems that the independence of B risks to be at the mercy of Yodel, after all, the company sets the standards and reserves the faculty to reject substitutes based on that.

The analysed elements do not appear to be indicia to discard subordination, and consequently an employment relationship in the case at hand. Instead, those elements are identified in the context of an EU Directive and case-law protecting workers as still consistent with an employment relationship.

Third, by making a reasoned order, the CJEU reinforces the development of the de facto precedent of the CJEU by stressing the importance of previous rulings. This idea has the potential to modify, in practice, the horizontal relationship between the CJEU and national courts. It also makes the relationship between the CJEU and national courts multilateral.[28] This would potentially bring controversial consequences in respect of future case law in this matter. In any case, the real impact of this case remains to be seen in the application by the Watford ET.

*Ricardo Buendia is a lawyer from the University of Chile (Chile). Buendia is currently a Ph.D. Candidate at the University of Bristol (United Kingdom), where he studies the idea of domination in the context of the gig-economy in Europe.

[1] Joined Cases C-297/88 & C-197/89, Massam Dzodzi v. Belgian State, 1990 E.C.R. I-3763, ¶ 38.
[2] Treaty on the Functioning of the European Union art. 267, Oct. 26, 2012, 2012 O.J. (C 326) 1.
[3] See Case C-35/76, Simmenthal v. Ministero delle Finanze, 1976 E.C.R. 1871.
[4] Rules of Procedure of the Court of Justice, 2012 O.J. (L 265).
[5] Paul Craig & Gráinne de Búrca, Preliminary Rulings, in EU Law: Text, Cases, and Materials 464, 478 (6th ed. 2015).
[6] Case C-692/19, B v. Yodel Delivery Network Ltd., 2020 ECLI:EU:C:2020:288 ¶¶ 21, 22.
[7] See Sandra Fredman & Darcy Du Toit, One Small Step Towards Decent Work: Uber v Aslam in the Court of Appeal, 48 Indus. L. J. 260 (2019); Nicola Countouris, Uses and Misuses of Mutuality of Obligations and the Autonomy of Labour Law, in The Autonomy of Labour Law (Alan Bogg et al. eds., 1st ed. 2015).
[8] See Nicola Kountouris, The Concept of “Worker” in European Labour Law: Fragmentation, Autonomy and Scope, 47 Indus. L. J. 192 (2018); Martin Risak & Thomas Dullinger, The Concept of “Worker” in EU Law (2018).
[9] B v. Yodel Delivery Network Ltd., supra note 6, at ¶ 24.
[10] Id. at ¶¶ 26,29; Kountouris, supra note 8.
[11] See Case C-413/13, FNV Kunsten Informatie en Media v. Staat der Nederlanden, 2014 ECLI:EU:C:2014:2411.
[12] See Craig & de Búrca, supra note 5.
[13] Case C-270/13, Iraklis Haralambidis v. Calogero Casilli, 2014 ECLI:EU:C:2014:2185.
[14] B v. Yodel Delivery Network Ltd., supra note 6, ¶ 45.
[15] Id. at ¶ 27.
[16] Id.
[17] Id. at ¶ 28.
[18] Id. at ¶ 27.
[19] ILO Recommendation R198: Employment Relationship Recommendation (95th session, Geneva, May 31, 2006).
[20] Id. at ¶ 9.
[21] See Sergio Gamonal C. & César F. Rosado Marzán, Chapter 3: Primacy of Reality, in Principled Labor Law: U.S. Labor Law through a Latin American Method Chap (2019).
[22] See Jeremias Prassl, Humans as a Service: The Promise and Perils of Work in the Gig Economy (2018); Adrian Todolí-Signes, The End of the Subordinate Worker? Collaborative Economy, On-Demand Economy, Gig Economy, and the Crowdworkers’ Need for Protection, 33 Int'l J. Comp. Lab. L. & Indus. Rel. 241 (2017).
[23] See Valerio De Stefano & Antonio Aloisi, European Legal Framework for “Digital Labour Platforms” (2018); Breen Creighton & Shae McCrystal, Who Is a “Worker” in International Law?, 33 Comp. Lab. L. & Pol. J. 691 (2016).
[24] B v. Yodel Delivery Network Ltd., supra note 6 ¶¶ 37, 43.
[25] European Parliament and Council Directive 2019/1152 2019 O.J. (L 186) (EU).
[26] See Bartłomiej Bednarowicz, Delivering on the European Pillar of Social Rights: The New Directive on Transparent and Predictable Working Conditions in the European Union, 48 Indus. L. J. 604 (2019).
[27] See Pimlico Plumbers Ltd v. Smith [2018] UKSC 29.
[28] See Craig & de Búrca, supra note 5.





Dispatch No. 23 - Russia - "Regulating Digital Online Work in the Russian Federation"

by Mariya Aleksynska*, Elena Gerasimova**, & Nikita Lyutov***[1]

May 2020


Since the late 1990s, the Russian Federation has been witnessing a proliferation of digital technologies. Initially low, access to Internet grew rapidly in the past two decades, with Internet users reaching, in 2017, 76,4% of the population (ITU, 2020). The broadband penetration rate is particularly high in big cities and in Central Russia. Even if there are important inequalities across the country in terms of access and the use of these technologies, the Russian-speaking Internet is now covering over 100 million people, and is the largest in Europe (ibid).

Naturally, many businesses integrated the Internet-related technologies and communication tools into their working processes. These technologies also modified the business processes and the work modes, in particular, by allowing workers to work remotely from home or other places outside of their regular workplace. New modes of work also appeared. Notably, since the mid-2000-ies, the Russian Federation experienced an emergence of a new type of work – digital work through online labor platforms and mobile apps. The Russian market for online platform work enjoyed a particular growth in the 2005-2009 period with the development of several general profile and specialized online platforms in various occupations (Strebkov, Shevchuk, Spirina, 2015). The growth of online work was further exacerbated by the global economic recession of 2008–2009, and a new economic recession of 2014. Already by 2014, at least 70 digital labor platforms for remote work operated in the Russian-speaking Internet (ibid). By 2019, at least one of them counted over 3 million active users (www.advego.ru: 3.250k in September 2019), and at least two counted over 1 million users (www.etxt.ru: 1.000k, www.fl.ru: +1.600k in September 2019). Russian market for online work features a very strong local dimension, as Russian-based workers are often matched with Russian-based clients. Moreover, according to various estimates, the Russian Federation appears among the leading economies in the world to match its workers with clients based internationally (Graham et al., 2017; OLI index, described in Kassi and Lehdonvirta, 2018). In addition to this, work through mobile apps, in which workers and clients are matched by a mobile app, but the work is executed locally, gained particular prominence since the mid 2010-es (Aleksynska, 2020).

To respond the development of internet technologies, and to the modifications that they bring to the existing modes of work, in 2013, the government of the Russian Federation introduced the new chapter in its Labour Code to include regulations of online work. As a matter of fact, because the phenomenon of digital work is still quite new world-wide, this regulation is among the very few currently officialized attempts to regulate Internet-mediated work in general (ibid). It is for this reason that this legal change is examined here with the aim of highlighting its innovativeness, but also some of its limitations.

In 2013, the Labor Code of the Russian Federation (further – the Labor Code) was amended by the Federal Law N.60-F31 to include Chapter 49.1 “On distinctive features of remote work of employees”[2] (Gerasimova, 2013; Filatkina, 2015; Lyutov and Tsypkina, 2018). The new chapter aimed specifically at legalizing and regulating “on-distance”, “remote” employment relationships that are mediated by digital technologies, and by Internet in particular. Article 312.1 of the Labor Code defines its scope of application as pertaining uniquely to work performed outside of employer’s premises and outside a standard workplace controllable by an employer, and for which ICT networks are used for communication between the employer and the employee on execution of a job. Remote employees are considered to be those who concluded an employment contract about remote work, and are covered by labor law, taking into account the particularities of this work. They should receive a salary, rather than civil-contract-type compensations, have a right to paid leave, and be considered as employees.[3] The control over employee is performed remotely, with electronically set milestones and deadlines and sent work by those dates considered as proofs of work being done (para. 1, art. 312.1 of the Labor Code). The remote employees thus defined are different from home workers, which have been also regulated by the Labor Code (chapter 49 of the Labor Code). Differences, among others, include the workplace: for a home worker, the workplace is the employer’s organization (para. 1, art. 310 of the Labor Code), while for a remote worker, the workplace is his or her own home, without territorial limits (para 1, art. 312.1 of the Labor Code). The list of works to be performed remotely, however, is not fixed by the legislator.

These new provisions also grant remote workers a number of exemptions from the general rules applicable to hiring and execution of employment contracts. For example, the parties are allowed to exchange documents in an electronic form (“electronic documents”) as long as they are authenticated with a so called “reinforced qualified electronic digital signature”, which in Russia are formal legal terms also defined by legislation. They are also allowed to abstain from maintaining “work booklets” - the official track record of a worker’s life-long experiences, which all employers and employees are obliged to keep. Finally, the law allows a remotely working employee to set his or her own work and rest regimes in a unilateral manner (Gerasimova, Korshunova, Chernyaeva, 2017), unless otherwise is provided by employment contract.

Other particularities include provisions on confidentiality and information protection clauses that should be part of the employment contracts of remote employees. They should concern the company information accessed by the remote worker, but also that is produced by the remote worker for the company. Thus, the law foresees that both remote worker and employer should follow the rules of information safety. Personal computers of remote employees can be equipped with crypto-protection systems, preventing information sharing with third parties. In return, a remote worker can ask for a purchase or amortization of computer equipment necessary to execute work (Lyutov and Tsypkina, 2018). Employers of remote employees also carry much less obligations on occupational safety and health comparing to regular employees and home workers (Gerasimova, 2013).

Despite these innovations, the new regulation is seen only as a first step towards creating a real regulatory field for online work. In fact, the concept of remote work does not cover remote work performed for an employer located outside of the Russian Federation. As a result, there are currently no provisions on the rights and obligations of workers operating remotely across borders, and neither there are provisions to guarantees them equal treatment or safety have been established (Gerasimova, Korshunova, Chernyaeva, 2017). Moreover, for the new provisions to be effective and complete, they have to be supplemented by a series of amendments, including a possibility to execute remote arrangement only on a part-time or temporary basis, otherwise combining it with regular work with the same employer (Gerasimova, Korshunova, Chernyaeva, 2017). They should also be complemented by the “right to disconnect” (Lyutov, 2018); more clear regulation of working time of remote workers; and a more clear set of criteria of proving the existence of an employment relationship (Lyutov, 2019). The subordination criteria, necessary for establishing an employment relationship, should also be complemented by the economic dependency criteria, as well as by the acknowledgment of the asymmetry in economic powers of both parties of the agreement (ibid). Some scholars also regret the abolishing of “work booklets”, suggesting that technologies allow to have “electronic work booklets” (Kurennoy, Kostyan, Khnykin, 2017), which may be a useful tool for proving work experience and for claiming social security (particularly pension) rights. A new law, dedicated to the issue of electronic booklets, is to enter in force in 2020; its efficacy is yet to be assessed.

The authors of this new Labour Code chapter had in mind that the new rules would cover primarily those remote employees whose employment relationship is mediated by information and communication technologies (ex.: IT specialists, accountants, translators, content writers, designers, journalists, lawyers, etc.). Yet in practice, these new norms are also frequently applied towards trade agents and other employees whose work requires regular travelling and who use ICT among other means of contacts.

In contrast, the vast majority of online platform workers, and workers of the mobile apps, continue working outside of the scope of this legal provision, despite the fact that their relationships with online platforms, or with online clients, often have all features of the remote employment relationship. Part of the reason for this is that the new provision did not aim to regulate platform work and work via apps specifically. Rather, it adopted a large definition of work mediated by ICT technologies, which primarily concerned those traditional employment relationships that were witnessing modifications under the influence of technologies. As such, the law did not set any specific incentives to convert contractual relationships on the online labor platforms and on mobile apps into the remote employment relationships. In a few rare cases when workers attempted to challenge their status in courts, asking for a reclassification, they did not succeed so far. Indeed, there were two attempts of “Yandex-Taxi” (Russian online platform for personal transportation services) drivers to challenge their lack of employment status in courts.[4] In both cases the courts simply referred to civil law ‘labels’ described in the contracts, and not to the new Labour Code provision, also ignoring the actual facts relating to the matter.

To sum up, the new Russian regulation can benefit from further improvements to be fully effective and to reach out the online platform workers, who are covered by civil contracts rather than by labor law, or work informally. At the same time, this regulation provides a formidable example of challenges that a national legislator may face when attempting to regulate online work. It also shows that an appropriate application of the law, its interpretation, and its enforcement, are key to an effective functioning of any law.  


*Mariya Aleksynska (IEDES, Paris I Pantheon Sorbonne, Paris)

**Elena Gerasimova (International Labour Office, Suva City)

***Nikita Lyutov (Kutafin Moscow State Law University, North West University, South Africa)


[1] The views expressed in this article are solely those of the authors. They do not necessarily reflect the views of their institutions.

[2] See Federal'nyy zakon o vnesenii izmeneniy v otdel'nyye zakonodatel'nyye akty Rossiyskoy Federatsii [Federal Law of the Russian Federation on Amending Certain Legislative Acts of the Russian Federation], Sobranie Zakonodatel’stva Rossiiskoi Federatsii [SZ RF] [Russian Federation Collection of Legislation] 2013, no. 14, Item 1668 [http://www.consultant.ru/cons/cgi/online.cgi?req=doc&base=LAW&n=321548&fld=134&dst=1000000001,0&rnd=0.4576204591208083#08327743398711546].

[3] See Nikita Lyutov & Elena Gerasimova,  The Concept of 'Employee': The Position in Russia, in Restatement of Labour Law in Europe 577 (Bernd Waas, Guus H. van Voss eds., 2017) (regarding the notion of employee in Russia).

[4] Resheniye Tushinskogo rayonnogo suda g. Moskvy ot 26 iyunya 2019 goda po delu no. 2-2238/19 po isku YA.I. Shcherbinina  k OOO “Yandeks Taksi” ostavleno bez izmeneniya Apellyatsionnym opredeleniyem Moskovskogo gorodskogo suda ot 22 noyabrya 2019 goda po delu no. 33-53437/2019; Resheniye Zamoskvoretskogo rayonnogo suda g. Moskvy ot 14 maya 2019 g. po delu no. 2-2792/2019 po isku V. YU. Golovanova k OOO “Yandeks Taksi.”



Mariya Aleksynska, Digital Work in Eastern Europe: Overview of Trends, Outcomes, and Policy Responses (2020) (forthcoming working paper) (International Labour Organization).

Elena Gerasimova, Remote and Home-Based Work: The Differences, no. 10 Kadrovik 16 (2013). In Russian: Елена Герасимова, Дистанционная и надомная работа: отличия, no. 10 Кадровик 16 (2013).

Elena Gerasimova et al., New Russian Legislation on Employment of Teleworkers: Comparative Assessment and Implications for Future Development, no. 2 Pravo. Zhurnal Vysshey Shkoly Ekonomiki, 116, 121 (2017).

Mark Graham et al., Digital Labour and Development: Impacts of Global Digital Labour Platforms and the Gig Economy on Worker Livelihoods, 23 Transfer 135 (2017).

Anna P. Filatkina, Homework and Remote Work: Comparative Analysis of Current Legislation, no. 1 Labour Law in Russia and Abroad 11 (2015).

Otto Kässi & Vili Lehdonvirta, Online Labour Index: Measuring the Online Gig Economy for Policy and Research, 137 Technological Forecasting and Social Change 241 (2018).

Alexander Kurennoy et al., Digital Economy of Russia. Electronic Administering of Labour Relations, no. 37 Electronic J. Lawyer 10 (2017). In Russian:  Александр Куренной и другие, Цифровая экономика России. Электронное делопроизводство трудовых отношений, no. 37 ЭЖ-Юрист 10 (2017).

Telecommunication/ICT Indicators Database, ITU (Dec. 2019), https://www.itu.int/en/ITU-D/Statistics/Pages/publications/wtid.aspx.

Nikita L. Lyutov, Remote Work: Experience of the European Union and Problems of the Legal Regulation in Russia. 143 Law Improvements 30 (2018). In Russian: Никита Л. Лютов, Дистанционный труд: опыт европейского союза и проблемы правового регулирования в Pоссии, 143 Cовершенствование законодательства 30 (2018).

Nikita L. Lyutov, Transformation of Employment Relationship and New Forms of Employment at the Times of Digital Economy 103 J. Russ. L. 98 (2019). In Russian: Никита Л. Лютов, Трансформация трудового правоотношения и новые формы занятости в условиях цифровой экономики, 103 Журнал российского права 98 (2019).

Denis Strebkov et al., Development of the Russian-Language Market for Remote Work, 2009–2014 (Based on the Results of the Census of Freelancers) (2015). Ed.: Radaev, V.. Vol. 16. Publisher: Higher School of Economics. In Russian: Денис Стребков и другие, Развитие русскоязычного рынка удаленной работы, 2009–2014 гг. (по результатам Переписи фрилансеров). Отв. ред.: В. В. Радаев. Вып. 16. М. : Издательский дом НИУ ВШЭ. (2015).








Dispatch No. 22 - Poland - "Polish Legislative Reform Tests a More Principled Approach to Collective Rights of Self-Employed Workers"

by Zuzanna Muskat-Gorska* 

April 2020


Until 2019 Poland had one of the highest levels of workers performing personal work outside employment relationship who were at the same time fully excluded from the right to form or join trade unions or to bargain collectively. This situation changed on January 1, 2019 when a ground-breaking revision of Polish collective labor law entered into force.[1] Following comments made by the supervisory bodies of the International Labour Organization and a ruling of the Polish Constitutional Tribunal, the Government extended the application of the whole Trade Union Act to both own-account non-employees[2] and their employers.[3] There are conditions of such an extension: to be covered, persons performing remunerated work outside an employment relationship should not employ other workers and should have “rights and interests related to performing work that can be represented and defended by a trade union.”[4]  Does this extension include collective bargaining rights of self-employed workers? That was the legislator’s intention.[5] But at the same time, there is still no clarity about how the new criteria will be interpreted and how the law will be applied to one-person businesses that may fall under the scope of antitrust restrictions under EU competition law.

I. Polish Labor Law - No Third Category, but Unlimited Use of Service Contracts
In principle, the Polish labor market is characterized by a binary segmentation between subordinate or dependent employment on the one hand and autonomous or independent self-employment on the other.[6] However, in practice, the Polish law offers numerous possibilities to engage dependent work based on a civil contract. The most typical examples of such contracts are a task-specific contract (umowa o dzielo) or a contract for the provision of unspecified tasks (contract of service, umowa zlecenia). These contracts can be offered both to persons without registered business activity as well as the “self-employed.” There are no particular safeguards against the use of civil contracts to engage labor or services. On the contrary, the Supreme Court has repeatedly recognized the legality of contracting labor based on a civil contract for any type of work[7] and stressed the importance of the will of the parties – and the right – to freely choose the legal basis of employment.[8] At the same time, judicial criteria of classification remain unclear, overlapping and somewhat circular – on the one hand, the employment relationship is defined by criteria such as personal, continuous character of work, remaining at the disposal of the employer in specified time and place under the employer’s supervision and management, on the other – none of these criteria, individually or cumulatively, are decisive for employment classification, since each of them, according to the courts, can be found in a relationship governed by a service contract.[9] Accordingly, performance of long-term, personal, continuous work, even in core activities of the employing activity, next to colleagues performing comparable work based on employment contract, while remaining at disposal of the employer can be still qualified as lawful engagement based on civil contract.[10] The Supreme Court has also repeatedly denied the existence of the presumption of an employment relationship under Polish law, albeit opinions to the contrary can be found in scholarship.[11] As a result of such liberal interpretation the Polish labor market is characterized by a high incidence of non-standard forms of work. It is estimated that at least 1 million workers without a registered business activity are employed solely on civil contracts and around 3 million workers operate as one-person businesses, but the numbers may be much higher.

II. Description of the New Legislation
Without any seniority requirement, and regardless how much of income they get from a given company and whether or not they have contracts with other companies, own-account non-employees – who “have rights and interests related to performing work that can be represented and defended by a trade union” – can now join enterprise-level trade union organizations and be represented by them both in individual as well as in collective matters. After six months of work for a given company (arguably including non-consecutive spells), trade union rights of non-employees expand. They can form a mixed or even a purely non-employee enterprise-level trade union organization. In many matters such organization has the same collective rights as an employee trade union, for example, in terms of the right to initiate and to take part in negotiations of a collective agreement[12], a collective dispute, or a strike.[13] These organizations have equal right to be granted premises and equipment necessary for the performance of their functions as employee unions.[14] Purely non-employee trade unions can also represent non-members who are employees. Some collective matters remain reserved to unions that represent at least 5% of employees. These matters include negotiation and conclusion of an agreement over enterprise-level rules concerning salary, working time or work organization.[15] Individual trade union rights of workers performing personal work outside of an employment relationship have also expanded. The law extends a 6-month employment protection – i.e. protection against early termination of the contract or unilateral modification of terms and conditions - to the non-employee union founders.[16] Non-employees can now be elected trade union officials, including by a union representing mainly employees. In such case, they are entitled to protection of their contracts against early termination or change of terms and conditions – but not to an extension of the contract if it ends before the end of mandate.[17] In case of unlawful termination of a contract with a protected non-employee no reinstatement claim is available (similarly as in the case of fixed-term employees) but the employer must pay compensation equal to 6 months of non-employee’s salary. Like employees, non-employee trade union leaders can be granted time-off facilities to carry out their functions in the amount of hours per month equal to the number of represented workers.[18] In cases concerning violation of these trade union rights, non-employees gain access to labor courts.

Overall, the scale of legislative change is impressive. The law does not expressly exclude any types of contractual arrangements from its scope. Nor does it refer to criteria more typically used to grant collective bargaining rights to self-employed workers – such as an element of personal subordination (dependency) combined with a quantitative threshold of income depending on a limited number of principals.[19] Instead, the law refers – without providing a definition – to the types of rights and interests that persons performing own-account work should have in order to be covered by the collective labor law. This seems like a potentially inclusive formulation, capable of catching many workers in non-standard forms of employment, including in platform economy. However, it remains to be seen how the law will be applied to collective bargaining covering self-employed workers who may fall under the scope of antitrust restrictions under EU competition law.

III. Dealing with the Threat of EU Competition Law
Self-employed workers, if viewed as independent entrepreneurs, are principally barred from engaging in collective bargaining about their remuneration under EU competition law (Article 101 TFEU).[20] According to the case law of the Court of Justice of the EU, an association representing self-employed workers[21] – or both employees and self-employed workers[22] – does not act as a trade union and therefore as a social partner, but acts as an association of undertakings.[23] An exception to that rule is only possible if the service providers in the name and on behalf of whom the trade union negotiated are in fact false self-employed, that is, service providers in a situation comparable (n.b. not “identical”)[24] to that of employees.[25] Accordingly, the current specific notion of workers for competition law, as forged by CJEU in the landmark 2014 case FNV Kunsten, continues to make reference to subordination as the criterion traditionally used to establish worker status in labor law. Subordinate employment relationship is characterized by the presence of direction of the employer[26], as regards, in particular, the employee’s freedom to choose the time, place and content of his work.[27] However, the Court also refers to more economic and organizational criteria, such as sharing of commercial risk[28], and being integrated into the organization of the employer’s undertaking, that is, forming an economic unit with that undertaking.[29] Lack of genuine undertaking status, i.e. false self-employment, is indicated in situations where a service provider does not determine independently his own conduct on the market, but is entirely dependent on his principal, because he does not bear any of the financial or commercial risks arising out of the latter’s activity and operates as an auxiliary within the principal’s undertaking.[30]

The notion of workers for competition law purposes, as forged by CJEU in FNV Kunsten, would arguably not be sufficient to cover all those in atypical forms of work who mainly trade their personal labor, and remain in situation of economic dependency on their employers.[31] However, this definition could already help distinguish categories of Polish self-employed workers to whom competition law should not apply.

First, contrary to the CJEU case-law,[32] the Polish law seems to accept that independence of a service provider may be merely notional, if so agreed by both parties to the contract. Accordingly, a “B2B” relationship may be regulated in a way that leaves little room for autonomy. The Polish contract of service is structured around personal performance[33] of specified – but not task-specific work. It entails obligation of due diligence, not of result. The service provider is obliged to follow binding instructions as regards performance of tasks.[34] Both time and place of performing work can be specified in the contract, as long as it is a free agreement of the parties.[35] Accordingly, the presence of subordination to the direction of the employer does not preclude the parties to classify their relationship as one of self-employment.[36] At the same time, the absence of elements of control precludes classifying a contract as an employment one.[37] It seems that the Polish test of “subordination” may be sometimes stricter than the test used in the CJEU’s case-law. The latter may accept weaker forms of subordination such as obligation to report and cooperate or being subjected to general, not detailed, direction or supervision, while enjoying a margin of discretion in the performance of duties.[38] Further, the business model of many Polish companies is built on the use of service providers whose work form an integral part of the company’s business. The law does not prevent a service provider to use the tools or equipment provided by the principal’s undertaking or to work only for one contractor. Remuneration of service providers can be hour-based, even if such arrangement precludes service provider’s control over working time or possibility to work for multiple contractors. In fact, tax authorities currently estimate that at least 0.5 million of solo firms are dependent on one contractor as regards at least 75% of income. Accordingly, Polish service providers, even when subject to weaker forms of subordination, especially in cases where they do not bear the employer’s commercial or financial risks (or where a contractual agreement shifts such risks to the service provider in a way that does not defy his or her worker status)[39] and form an integral part of the employer’s undertaking, could be found to be in a situation comparable to that of employees and therefore having full collective bargaining rights.

The new Polish law lands in the middle of regulatory debate concerning access of non-standard workers – and the self-employed workers in particular – to fundamental labor rights to organize and bargain collectively. In the context of EU law it is becoming apparent that the notion of worker forged in FNV Kunsten may not be sufficient to eliminate conflicts between labor law and competition law and uncertainty as to their respective scope. Against this backdrop came the words of the EU Commissioner for Competition Margrethe Vestager who in October 2019 publicly defended the right of on-demand workers to unionize and fight for proper wages.[40] “We need to make sure that there is nothing in the competition rules to stop those platform workers from forming a union, to negotiate proper wages as you would do in any other business,” Vestager said.[41]

The fact that the new Polish law does not refer to any of the typical criteria used to grant collective bargaining rights to self-employed workers may be one reason for international scholars to show an interest in this legislative reform. Such model of regulation creates an opportunity to develop a more up-to-date doctrine on the relation of competition law and collective bargaining, that goes beyond FNV Kunsten and takes account of the principle and objective of the “social market economy” which is common to the Polish[42] and EU’s[43] constitutional orders. Another reason is the fact that the new Polish law goes far beyond giving trade unions the right to represent non-standard workers. Giving only right to join unions does not fully respect the entitlement of non-standard workers to collective voice. Those hired outside of employment relationship may have particular concerns they would like to address through trade union activity. For example, they may be “more concerned with wage rates or pathways to permanent employment than promotion hierarchies.”[44] Unfortunately, it is still not clear how the Polish legal system intends to deal with application of the amended Trade Union Law to own-account entrepreneurs. The worst scenario would be that each such worker would have to demonstrate that they may be covered by terms of a collective agreement or take part in collective action, a procedure requiring time and effort.[45] It remains to be seen how the law will be interpreted and applied in practice.

*Zuzanna Muskat-Gorska: External Scientific Collaborator at the at the Institute for Labour Law, Katholieke Universiteit Leuven, Belgium.


[1] Labour Code of 26 June 1974 (Dziennik Ustaw., 1974 Nr 24 poz. 141), as amended by the Law of July 5 , 2018 (entered into force on January 1, 2019).

[2] Id. art. 11(1) and art 2.1.

[3] Id. art. 11(2).

[4] See OECD, Policy Responses to New Forms of Work 1, 73 (2019), available at http://www.oecd.org/employment/policy-responses-to-new-forms-of-work-0763f1b7-en.htm (description of the legislative change reported by the Polish Government).

[5] Id. 

[6] See Leszek Mitrus, New Forms of Employment in Poland, in New Forms of Employment in Europe 305, 306  (Roger Blanpain et al., eds. 2015)2; Andrzej Marian Świątkowski & Marcin Wujczyk, The Classification of Employment in Polish Labour Law, 27 K. L. J. 165 (2016).

[7]  See argumentation by the Appeal Court upheld in II PK 372/12 of September 11, 2013 of the Supreme Court.

[8] Id.

[9] See id. See e.g., For critical overview of criteria of the employment test in the Polish law see, e.g.,, Michal Pawłucki, Kierownictwo Pracodawcy – Cecha Konstrukcyjna Stosunku Pracy Dezorganizująca System Form Zatrudnienia [Employer Management – A Feature Construction Relationship Disorganizing the System of Employment Forms], 10 Internetowy Przegląd Prawniczy TBSP UJ 56 (2017) .

[10] See II PK 372/12 of September 11, 2013 of the Supreme Court.

[11]   Andrzej Marian Świątkowski, Prawo Wyboru Podstawy Zatrudnienia [The Right to Choose a Legal Source of an Employment Relationship], 24 Studia Iuridica Lublinensia 237 (2015) (Poland).

[12] Art. 24124-24125, 24128 k.k. (Dz. U 1974 Nr 24 poz. 141) as amended by the Law of July 5, 2018.

[13] Art. 6, Act of 23 May 1991 on the Settlement of Collective Labour Disputes (1991 r. DZ. U. Nr 55 poz. 732); Art. 11(1), Trade Unions Act of 23 May 1991, as amended by the Law of July 5, 2018 (1991 r. DZ. U. Nr 55 poz. 725).

[14] Art. 33 Art. 33 and 26, Trade Unions Act of 23 May 1991, as amended by the Law of July 5, 2018 (1991 r. DZ. U. Nr 55 poz. 725).

[15] Id. art. 30.6.

[16] Id. art. 32.

[17] Id. art. 32.2.

[18] Id. art 31.1.

[19] See e.g., Lionel Fulton, Trade Unions Protecting Self-Employed Workers (2018) (outlining a comparative review of laws extending collective bargaining rights to self-employed workers in the EU); Nicola Countouris & Valerio De Stefano, New Trade Union Strategies for New Forms of Employment (2019). 

[20] Countouris & De Stefano, supra note 18; Dagmar Schlek & Andrea Gideon, Outsmarting the Gig-Economy Through Collective Bargaining – EU Competition Law as a Barrier to Smart Cities?, 32 Int’l Rev. L. Computers & Tech. (2018); Ioannis Lianos et al., Re-thinking the Competition Law/Labour Law Interaction: Promoting a Fairer Labour Market, 10 Eur. Lab. L. J. 291(2019). .

[21]  See Case C-180-184/98, Pavlov v. Stichting Pensioenfonds Medische Specialisten, 2000 E.C.R. I-06451; Case C-309/99, Wouters v. Algemene Raad van de Nederlandse Orde van Advocaten, 2002 E.C.R. I-01577.

[22] See, e.g., Case C-413/13, FNV Kunsten Informatie en Media v. Staat der Nederlanden, 2014 E.C.R. 2411.

[23] Id., para 28.

[24] See  Countouris &  De Stefano, supra note 18, at 50.

[25] Case C-413/13, FNV Kunsten Informatie en Media v. Staat der Nederlanden, 2014 E.C.R. 2411, paras. 30-31.

[26] Id., para 34. For a more detailed discussion of the concept of employee in EU law see Nicola Countouris, The Concept of ‘Worker’ in European Labour Law: Fragmentation, Autonomy and Scope, 47 Indus. L.J, 192 (2017).

[27] Case C-413/13, FNV Kunsten Informatie en Media v. Staat der Nederlanden, 2014 E.C.R. 2411, para 36, (referring to Case C-256/01, Allonby v. Accrington & Rossendale College and others, 2004 E.C.R. I-00873, para 72).

[28] Id. (referring to judgement in Case C-3/87, The Queen v. Ministry of Agriculture, Fisheries, and Food, ex parte Agegate Ltd., 1989 E.C.R. 04459, para 36).

[29]Id. (referring to Case C-22/98, Criminal Proceedings Against Becu, 1999 E.C.R. I-05665, para 26).

[30]Id. para 33 ( referring to Case C-217/05, Confederación Española de Empresarios v Compañía Española de Petróleos SA, 2006 E.C.R. I-11987, paras 43-44).

[31] See e.g., Schiek & Gideon, supra note 19; Countouris & De Stefano, supra note 18; Lianos et al., supra note 19. 

[32] Case C-413/13, FNV Kunsten Informatie en Media v. Staat der Nederlanden, 2014 E.C.R. 2411, para 35, (referring to judgment in  Case C‑256/01, Allonby v. Accrington and Rossendale College, 2004 E.C.R. I-00873, para 71).

[33] Unless otherwise specified in the contract.

[34] Kodeks cywilny (k.c.) [Polish Civil Code 1964] (1964 r. Dz.U. Nr 16, poz. 93) art. 737. See also Edward Gniewek & Piotr Machnikowski, Kodeks Cywilny Komentarz [Civil Code Comment]  (7th ed. 2016) (commenting on art. 737).  

[35] See I PKN 389/98 of October 6, 1998 of the Supreme Court.

[36] Pawłucki, supra note 8.  See also II PK 372/12 of September 11, 2013 of the Supreme Court; II PK 153/14 of April 22, 2015 of the Supreme Court.

[37] Pawłucki, supra note 8. See also I PKN 256/00 of February 14, 2001 of the Supreme Court.

[38] Case C-232/09, Dita Danosa v. LKB Līzings SIA., 2010 E.C.R. I-11450, paras 49, 51, 56.

[39] For this line of argument, see Schiek & Gideon, supra note 19, at 13; Case 3/87, The Queen v Ministry of Agriculture, Fisheries and Food, ex parte Agegate Ltd., 1989 E.C.R. 04459, paras 36-37.

[40] See Javier Espinoza, Vestager Says Gig Economy Workers Should ‘Team Up’ On Wages, Financial Times (Oct. 24, 2019), available at  https://www.ft.com/content/0cafd442-f673-11e9-9ef3-eca8fc8f2d65.

[41] Id. 

[42] “A social market economy, based on the freedom of economic activity, private ownership, and solidarity, dialogue and cooperation between social partners, shall be the basis of the economic system of the Republic of Poland”. Konstytucji Rzeczypospolitej Polskiej [Constitution] Dz. U. 1997, NR 78 poz. 483, art. 20 (Pol.).

[43] Labour Code of 26 June 1974, supra note 1, art. 3(3) (referring to realizing a highly competitive social market economy).

[44] Bettina Heidinger, Organizing Peripheral Workers in Parcel Delivery and Postal Services 199, 210, in The Outsourcing Challenge: Organizing Workers Across Fragmented Production Networks (Jan Drahokoupil ed., 2015).

[45] Id. at 213; Tonia Novitz, Changes in Employment Status Under Austerity and Beyond – Implications for Freedom of Association, 39 D. U. L. J. 27 (2016).


1. Michal Baranski & Iwona Gredka-Libanska, O Strukturze Związków Zawodowych po Nowelizacji Prawa Związkowego z Dnia 5 Lipca 2018 r [On the Structure of Trade Unions After the Amendment to the Trade Union Law of 5 July 2018], 78 Studia Iuridica 24, 24-39 (2018) (Poland).
2. Nicola Countouris, The Concept of ‘Worker’ in European Labour Law: Fragmentation, Autonomy and Scope, 47 Indus. L.J, 192 (2017).
3. Nicola Countouris & Valerio De Stefano, New Trade Union Strategies for New Forms of Employment (2019).
4. Lionel Fulton, Trade Unions Protecting Self-employed Workers (2018).
5. Mark Freeland & Nicola Countouris, The Legal Construction of Personal Work Relations (2011).
6. Edward Gniewek & Piotr Machnikowski, Kodeks Cywilny Komentarz [Civil Code Comment] (7th ed. 2016).
7. Bettina Heidinger, Organizing Peripheral Workers in Parcel Delivery and Postal Services 199, in The Outsourcing Challenge: Organizing Workers Across Fragmented Production Networks (Jan Drahokoupil ed., 2015).
8. ILO, Freedom of Association - Compilation of Decisions of the Committee on Freedom of Association (6th ed. 2018).
9. Ioannis Lianos, Nicola Countouris & Valerio De Stefano, Re-thinking the Competition Law/Labour Law Interaction: Promoting a Fairer Labour Market, 10 Eur. Lab. L.J. 291, 291-333 (2019).
10. Agata Ludera-Ruszel, Samozatrudnienie Ekonomicznie Zależne a Konstytucyjna Zasada Ochrony Pracy [Dependent Self-employment in the Light of the Constitutional Principle of the Protection of Work], 27 Roczniki Nauk Prawnych 43 (2017) (Poland), available at http://dx.doi.org/10.18290/rnp.2017.27.1-3.
11. Leszek Mitrus, New Forms of Employment in Poland, in New Forms of Employment in Europe 305 (Roger Blanpain et al., eds. 2015).
12. Tonia Novitz, Changes in Employment Status Under Austerity and Beyond – Implications for Freedom of Association, 39 D. U. L. J. 27 (2016).
13. OECD, Policy Responses to New Forms of Work (2019), available at http://www.oecd.org/employment/policy-responses-to-new-forms-of-work-0763f1b7-en.htm.
14. Michal Pawłucki, Kierownictwo Pracodawcy – Cecha Konstrukcyjna Stosunku Pracy Dezorganizująca System Form Zatrudnienia [Employer Management – A Feature Construction Relationship Disorganizing the System of Employment Forms], 10 Internetowy Przegląd Prawniczy TBSP UJ 56 (2017).
15. Adalberto Perulli, Subordinate, Autonomous and Economically Dependent Work: A Comparative Analysis of Selected European Countries, in The Employment Relationship: A Comparative Overview 137 (Giuseppe Casale ed., 2011).
16. Dagmar Schiek & Andrea Gideon, Outsmarting the Gig Economy Through Collective Bargaining – EU Competition As a Barrier to Small Cities?, 7 Intl. Rev. L. Computer & Tech. 275 (2018).
17. Andrzej Marian Świątkowski, Prawo Wyboru Podstawy Zatrudnienia [The Right to Choose a Legal Source of an Employment Relationship], 24 Studia Iuridica Lublinensia 237 (2015) (Poland).
18. Andrzej Marian Świątkowski & Marcin Wujczyk, (2016) The Classification of Employment in Polish Labour Law, 27 K.L.J 165 (2016).




Dispatch No. 21 - France - "The making of a French legal model of bullying: An empirical analysis of the France Telecom case"

by Gaëlle Deharo* 

February 2020





On May 6, 2019, the trial of the French company “France Telecom” began. This was the first time for a large CAC 40 index corporation to face criminal charges of “harcèlement moral.” This case is especially important because in the past decade at least 35 employees committed suicide,[1] and 12 people attempted suicide between 2007 and 2010.[2] There could be no doubt that 19 of those tragedies were connected to management practices, since the victims left letters referring directly to the management methods, or committed suicide at the workplace.

This case is extremely important as it pertains to the management system, from the bottom to the top. Seven people appeared before the French criminal court: France Telecom as a legal person, the CEO, the CEO’s number two and the head of human resources were prosecuted for bullying, and three middle managers were prosecuted for complicity in bullying. Even though France has specialized labor courts that make decisions based on special texts and special procedures, this case was brought before the criminal court and thus the criminal judge took on the mission of evaluating the management practices under the law.

It is significant to note that during the hearings the parties argued, on the one hand, the legitimacy of the managerial practices and, on the other hand, the basics of the law. This highlights the large gap between the technical managerial explanations of the accused, and the empathetic humane dimension of all the other parties (victims, relatives, employees, the occupational doctor, trade unionists, etc.).

From a scientific point of view, the trial provides a particularly interesting case for an inter-disciplinary analysis. The suffering of the employees, the explanations provided by the managers, the company’s organizational methods, and also the process by which individuals were broken down were laid bare at the hearings. This is an extraordinary field of research: hours of verbatim evidence are available to the researcher wishing to understand the processes that link law to business administration theory in the legal decision-making process.

From an empirical point of view, the case is all about action plans. On February 14, 2006, the CEO announced an acceleration of the company’s transformation plans, with the aim of implementing a “convergence of jobs.” More specifically, the case was all about the application of two action plans, named “Next” and “Act”, which implemented the entrepreneurial strategy of the reorganization of the company and explicitly provided for the loss of 22,000 jobs. From a managerial point of view, the leaders expected to save the firm in the complex and hostile economic context of global competition, European regulation and financial pressures. The company’s strategy thus responded to purely managerial and dehumanized concerns. Therein lies the problem: focused on the economic goals, the managers tried to adapt the organization to its entrepreneurial needs, regardless of the human beings involved.[3] It has to be noted that French law has a special procedure for mass redundancies, based on social plans. However, in the France Telecom case the company and the executives introduced a policy aimed at destabilizing employees in order to encourage them to quit.[4] This led to death, suicides, suicide attempts, broken families, destroyed employees, and shocks. The prosecutor requested the highest penalty provided by the law: one year’s imprisonment and a fine of €15,000, and a fine of €75,000 for the company. Should such a conviction be handed, the accused will also have to compensate the victims for all their damages.

In these circumstances, the judge was asked to assess not a strategy, but the methods followed to apply a strategy. Thus, the issue is: where does the border between management freedom and legal misconduct lie?

It is the purpose of this study to evaluate the theoretical and practical aspects of the France Telecom case, in particular as this case pertains to measurable impacts of business administration theories in the legal decision-making process. More specifically, the study aims to perform an analysis of the case on the basis of the linguistic formulas used at the hearings, in order to evaluate the role of managerial issues in the legal rationale. The methodology is founded on the points raised by the respective parties, in order to observe the use of business administration in the arguments. The key research question is: does business administration have any possible way to exonerate the accused persons from the liability that may result from their management methods?

The study should therefore determine where the borders of management freedom lie, and where personal dominance over employees begins. Indeed, French jurisprudence has specified that it is not abnormal to feel stress or pressure at work, but that it becomes abnormal as soon as the stress or pressure is too intense or is excessive compared to the tasks of the victim.[5] In other words, the French judge has to enter into the world of the company with their own tools: legal texts, jurisprudence and legal doctrinal studies. However, even in this perspective, 'there is a lack of accuracy in the notion of “harcèlement moral” in French law. In the France Telecom case, the textual gap has been filled by the accused, who argue that management concerns may be effective to exempt them from all liability. This is what the judge has to struggle with (I). Thus, there is a high risk that the application of the law will be governed by management imperatives (II). 

I. The legal definition of bullying in French law

The question of the definition of bullying is not purely academic. It has led to litigation that is significant from both a quantitative and a qualitative point of view. The French governmental website Légifrance,[6] which lists and makes available court decisions, reports 4,246 decisions between 2002 and 2019 devoted to bullying,[7] including 2,965 pronounced by the Cour de cassation, the highest court of the French legal system. Of these cases, 395 were heard in the criminal division and 2,460 in the social division. It is certainly easier to bring harassment proceedings before the civil courts, since the victim then benefits from a much more protective procedure: it is sufficient for him to report evidence suggesting the existence of bullying, and the person prosecuted must then prove his innocence.[8] The present case was brought in the criminal division. Faced with the human tragedy and the difficulty encountered by the victims in acting, it was a union that filed the criminal complaint. This means that all the parties had to demonstrate that the facts matched all the legal requirements, even though we do not know exactly what those requirements are about.[9] The first hypothesis in this study is that the notional ambiguity of bullying under the French rules increases the exposure of both managers[10] and employees to bullying.[11] In this respect, it should be remembered that French criminal law is founded on the logic that an act must qualify as a crime, based on the principle of legality: the existence of a text that makes the action criminal (A), but also the jurisprudential decisions on that text (B). 

A. The legal basis of the definition

Studies on the psychodynamics of work,[12] as well as studies in management, law, and psychology, highlight the essential importance of the employee’s feelings in the definition of bullying.[13] The way people feel about the management methods is very important in the process of describing those methods, but it would not be sufficient to claim that a malaise characterizes bullying.[14] The competitive environment of the employment may justify the fact that people are normally stressed. Thus, the purpose of this part is to identify the tipping point when lawful and legitimate management decisions, possibly felt badly by employees, become unlawful and illegitimate, and characterize the act of bullying.

In the France Telecom case, the accused were prosecuted for bullying because of the brutal management methods. When the case was being argued, the defense attorneys claimed that a criminal judge is not competent to give opinions about management decisions, thus referring to management freedom.

The term “bullying” cannot be easily defined.[15]  It is often used by people to express pain, but the term is basically not professional a term even if it may be applied in an employment context. Nevertheless, as a result of the work of a French psychiatrist, the characterization of “bullying” eventually appeared in the public debate in France.[16] The French legislator suddenly observed that people at work may suffer pain that they cannot make public because they are scared of losing their jobs. It was only in 2002 that the idea of bullying was codified in the Labor and Criminal Codes.[17] Even though the texts are very similar, two different procedures are available for pursuing culprits. The first is a civil procedure, brought before the labor court and only requiring that the victim makes a claim. In return, the defendant has to prove that the situation does not constitute bullying. The other procedure is a criminal one, brought before a criminal court and following criminal procedures.

In the France Telecom case, the procedure started in 2009 with a criminal complaint denouncing the harsh, brutal and pathological management practices. This led to French criminal processes being applied. One of the most important rules in French criminal law is the “principle of legality.” This is specified in article 111-3 of the French Criminal Code: “no one may be punished for a felony or for a misdemeanor whose ingredients are not defined by statute, nor for a petty offence whose ingredients are not defined by a regulation. No one may be punished by a penalty that is not provided for by the statute, if the offence is a felony or a misdemeanor, or by a regulation, if the offence is a petty offence.”[18] The article that follows specifies that criminal legislation is to be construed strictly.[19]

The principle of legality leads to three features being identified: the legal element, the material element and the intention to carry out harassment.

  • 1) Based on the written system, French law requires what is called the “legal element,” which refers to the existence of a text that makes the acts criminal offenses. The text in this case is article 222-33-2 of the French Criminal Code: “harassing another person by repeated conduct which is designed to or which leads to a deterioration of one’s conditions of work liable to harm  their rights and dignity, to damage physical or mental health or compromise career prospects is punished by a year’s imprisonment and a fine of €15,000.”
  • 2) The “material element” is defined by the text. In other words, the judge should refer to the legal requirements for the criminal act, match the legal requirements with the facts, and make a decision. In fact, article 222-33-1 of the French Criminal Code lays out cumulative legal requirements that characterize a situation of bullying in a professional context:[20]
  • - The first legal requirement is unequivocal: bullying is founded on the repetition of acts.[21] However, no explanation is provided by the text: on what kind of acts could bullying be based? In prior decisions, it has been held that the issue should be left for the judges to decide. The court has specified that it is not necessary that acts of the same kind are performed; the acts may be the same, or they may be different. The period of time does not matter either: the acts may take place over a very short period, or over a long one. Finally, the text does not refer to the status of “employee”, meaning that an employee might be prosecuted for the bullying of a manager.
  • - The second legal requirement offers an alternative: either the acts are designed to lead to a deterioration of the working conditions, or such a deterioration is a consequence of the repeated acts. The labor division of the French Cour de cassation[22] has specified that, because of this, there is no intention required to constitute bullying.[23] In other words, if prosecuted on the basis of bullying, a manager cannot usefully argue that there was no intention to carry out any bullying. Likely it would be inefficient to argue that there was no expectation to provoke such consequence.
  • - The third legal requirement specifies the consequences of the deterioration:
  • This deterioration harms the rights and the dignity of the victim, or
  • This deterioration damages the physical or mental health of the employee, or
  • This deterioration compromises the employee’s career prospects. In this case, the proof is quite complicated, since it has to be argued on the basis of speculation about the future.

It has to be remembered that each of these requirements represents a possible effect of the repeated acts, but the act may be characterized as bullying even if none of them occurs.[24]

  • 3) The “intentional element” refers to article 121-3 of the French Penal Code: “there is no felony or misdemeanor in the absence of an intention to commit it.” This means that the French judge could not convict someone without providing proof of an intention to carry out harassment. Therein lies a problem: it would be quite impossible to convict someone in most cases since there would be no explicit intention to carry out bullying. The text would therefore not be effective, and could not lead to a punishment for the bullying of a human being. That is why the criminal division of the French Cour de cassation specified that bullying may be present even if there is no intention, thus creating a special argument for bullying based on the special relationship between management and law in the legal decision-making process.[25]

B. The special relationship between management and law consecrated by French jurisprudence

The magistrate is in charge of the interpretation and application of the rules. That is why the judge is responsible for the definition of the acts that may characterize bullying. The France Telecom case had very particular arguments, since the accused did not refute the acts but tried to justify them as being part of the entrepreneurial context, the management objectives, and the business strategy. Thus, they asserted that all the acts were legitimized by the power to steer the company.

In this regard, this part aims to observe how the judge struggled with employment context in order to comply with the strict interpretation of article 222-33-1.

Under French labor law, the steering power enables the employer to take all the decisions necessary for the smooth running of the enterprise. However, this power is not absolute, and when exercising it the employer must respect the rights of the employees and ensure their safety. Prior decisions of the labor division of the French Cour de cassation established a link between bullying and management practices. According to those decisions, the management practices executed by a manager may constitute bullying, since an employee could identify repeated conduct that was designed to lead to or that led to a deterioration of their conditions of work  and the deterioration was liable to harm rights and dignity, to damage physical or mental health, or to compromise career prospects.[26]

More generally, the French Cour de cassation made a decision on June 15, 2017 in which for the first time it used the terms “managerial harassment” instead of “bullying.”[27] This description was already in use by researchers in management[28] and also in law.[29] The term “managerial harassment” appears more often in the decisions of the courts of appeal than in the decisions of the Cour de cassation. It refers to the application of harsh, brutal or oppressive management practices exceeding the limits of the normal steering power.[30] The prosecutor in the France Telecom case explicitly requested that “managerial harassment” be punished.

Indeed, the prior decisions of the criminal division of the Cour de cassation had already decided that bullying is based on behavior that exceeds the normal performance of rights or freedoms.[31] More specifically, it had been judged that a court of appeal acted properly when it characterized bullying and all its elements (legal, intentional and material) by highlighting facts that exceeded the “normal performance of the steering power” and harmed an employee’s dignity and career prospects.[32] In the same way, bullying is characterized as soon as it is found that the facts of which complaint is made are inherently not exercised under the steering power.[33] By contrast, there is no harassment as soon as the events of which complaint is made can be justified as part of the normal performance of the steering power.[34]

As a result, from the French point of view, “managerial harassment” occurs as soon as a system goes beyond the limits of the steering power, harshly, brutally, and oppressively. The France Telecom case was the first case brought to the court in which it was asked to evaluate a management system defined at the highest level of the firm. Obviously, the aim of all the management practices reported to the criminal court in that case was to reach a quantitative objective of 22,000 redundancies. The investigating judge specified that the case was not about the strategy, but was all about the management practices. In other words, the global strategy may be permissible but the methods may be questionable.

II. The emergence of an impact of management on the legal decision-making process?

French jurisprudence had already opened the path to assessing management methods. Some prior decisions had characterized certain management practices as “unsuitable,”[35] “brutal,” or “maladjusted.”[36] Moreover, the labor division of the Supreme Court enforces a proactive control by the lower courts when they have to check whether the management practices executed by a company could be characterized as bullying.[37] This means that the jurisprudence does not usually require harshly oppressive or degrading treatment: unsuitable management practices may constitute bullying.[38]

As a consequence, this is the question: can accused people, in return, argue that they were performing suitable management practices, and thus prevent a conviction based on bullying? This is the way the defense strategy seems to be designed in the case (A). However, focusing on the management practices within this single organization may lead us to miss the much larger picture: should management practices be taken into account by the judge (B)?    

A. Could suitable management practices exempt the accused from all liability?

The France Telecom case is symptomatic of a change in working relationships in France. In general, the implementation of a managerial type of rationality seems to be essential in the reasoning of both litigants and judges. The literature has highlighted that the choice of whether or not to go to trial, and the definition of judicial strategies, is indeed based on an objective that is no longer exclusively judicial but is also economic. In the case of France Telecom, the influence of  business administration theories on the law seems to be stronger. Pressure was exerted directly on the judge’s reasoning. Managers and a judge have a different understanding of how the business operates. The imposition of the managers’ own operating logic on the judge can be used as a basis for determining liability, so that liability is determined no longer on the basis of legal rules, but on the basis of management principles. Invoked by the defense, good management may become a justification for the behavior of leaders. At first sight, such a development is not to be feared, as the law and management have different fields of action. However, the contemporary evolution of labor relations questions the relationship between the law and managerialism.

The conception of the human being at work appears, in fact, as the paragon of a particularly important evolution in the articulation of the law and business administration. Most often, we observe that the organization is considered as a combination of technical and human factors.[39] The human being must be as efficient as the machine. As a result, under the influence of numerous studies in all areas of management that have highlighted the positive impact of well-being at work on productivity, organizations demand the well-being and even the happiness of their employees: employees are declared to be happy, managers to be benevolent. To achieve this, managers participate in “benevolent management” training, while employees are asked to learn how to manage stress or conflict, how to organize themselves better, and how to manage their work-life balance. There is an emerging demand for well-being and happiness for employees and, conversely, benevolence or empathy for managers.[40]   

From a legal point of view, this evolution raises questions because it tends to shift the focal point of the organization towards the employee: it is no longer the organization that is at fault, but the employee who is “fragile.” A quick observation of the practice of top and middle managers in French companies also sends employees back to their fragility and the poor management of their emotions, as if this were enough to close off definitively the question of the organization’s fault. It is from within this movement that France Telecom’s top management developed its defense strategy.

The defense discourse in this case is no longer only a tool of managerial legitimization, but it is also an argument that tends to exonerate the accused from civil and criminal liability. The exchanges between the lawyer for the civil parties and the former CEO of France Telecom, Didier Lombard, at the beginning of the trial are particularly significant in this regard. “I want to say to those who were my colleagues my deep sorrow for those who did not support the change imposed on the company,” says Didier Lombard. “Our “house”, in 2005, was in peril. Because of its over-indebtedness, the aggressiveness of the competition, technological developments (...) In 2009, France Telecom was doing better. And these results have been achieved thanks to the employees: I want to express my admiration and gratitude to them.” The situation, he says, “may have unintentionally contributed to weakening some, to the point that they perform an irremediable act, which is unbearable to me.” The lawyer asks: “Was there a social crisis at France Telecom?” The CEO’s answer is unequivocal: “No. There were difficulties. The transformations imposed on the company were not pleasant: OK. That’s the way it is. I can’t help it. If I hadn't been there, it would have been the same, maybe worse.”   

This argument is not insignificant. It clearly shows the desire to shift the focus from the law to management: it is not a question of “crisis” or even “bullying”, but a managerial necessity that has been misunderstood by employees. Moreover, Didier Lombard forcefully states that: “it is still how we saved the house.” His number two, Louis-Pierre Wénès, argued in the same vein. His arguments are intended to demonstrate that everything that has been done by the management team was justified by the managerial objectives. Economic success even seems to take precedence over the suffering of employees: “We deny everything that has been done for three years. We got the company out of the rut. I am furious that we are only looking at one side of the situation.[41]

Through managerialism management sciences, the employees’ suffering seems to become understandable and even justifiable: it is no longer the organization that is at fault, but it is the employee who is too fragile and cannot cope with the environment in the company. In this sense, one of the accused noted at the hearing that the people who committed suicide were “fragile.” Moreover, this is an explanation that is very often found in analyses and hearing reports. It must be strongly refuted, as the French psychiatrist Sylvaine Perragin does: there are no fragile people, she says. Burn-out and suicide are not consequences of fragility. People who suffer burn-out or commit suicide are people who are pushed to their physical and psychological limits by the tasks they are asked to achieve and are not given the tool to do those tasks.[42] In other words, suicide and burn-out are founded on the confrontation between a willingness to do the job well and the impossibility of doing the job well.[43] 

The argument of the accused rests on the idea that managerial justification would exonerate them from liability. By modifying the interpretation grid, the objectives become different: it is no longer a question of whether the accused is responsibility, but a question of whether the behavior was justified from a managerial point of view. Should we therefore see in the France Telecom case the emergence of the influence of business administration theories on the judge’s decision-making process?

B. Should the judge take management justification into account?

In reality, the question goes beyond France Telecom’s own framework. At the same time, several other companies were experiencing waves of suicides: Renault, La Post, EDF, and even the French national police.[44] All these organizations have in common that they have undergone a structural change from a public enterprise model to an entrepreneurial model. They also have in common that they are based on a very strong business approach.[45] However, the literature has identified the importance of each of these factors in the definition of bullying.[46] The hypothesis studied here is therefore that there may be a major, dangerous and worrying paradigmatic shift: must the judge assess the management methods by the economic goal they pursue? This was the question asked directly at the France Telecom trial. The hypothesis also questions a certain trivialization of managerial justification. So, to what extent should the judge take into account managerialism?[47] There is a significant discrepancy between the two sets of logic that the judge has to deal with in the France Telecom trial. To his normative logic, the judge must add a managerial dimension, based on the free choice of management methods to govern the smooth running of the company. The definition of management methods is part of the employer’s management power, with which the judge cannot interfere. Nevertheless, when assessing managerial methods, the judge does not enter a lawless territory. Under French law, the employer has many obligations towards his employees. In particular, the employer must provide the means to accomplish the employees’ tasks and ensure physical and mental safety. These obligations are sufficient to give a legal dimension to the choice of management methods. As a result, management methods are not a purely managerial concept but also govern the application of the law of employer liability. Although there is no legal definition of management or direction methods, the Cour de cassation does not hesitate to sanction them if it considers that the managers may be at fault. It is in this context that the judge takes into consideration the “ways of doing” that directly influence well-being and health in the company. 

By relying on the managerial relevance of their choices, the accused attempted to overturn this perspective and balance of logic in the judge’s decision-making process. This way of doing things is not purely French and can be observed, in particular, in large international groups. Under the guise of a social responsibility policy, companies impose happiness as a contractual obligation.

This change of perspective moves the pointer from the fault to the employee. However, it is not the employee who is unfit for happiness or well-being: it is the organization that does not give him the means to have happiness and well-being, by not allowing him to perform his task properly. It is also the judge’s responsibility to ensure that the employer evaluates the consequences of his managerial choices and the impact they may have on employees, in terms of psychosocial risks.

Clearly, management science has an effect on the application of the law. The development of the arguments of the defense highlights the porosity of both dimensions. However, it is above all the risk of a shift in the law under the influence of management that emerges: the assessment of fault would then be subordinated to the lack of managerial relevance of the choice made by the manager ... who is free to choose the same methods. In other words, it is the managerialism that is integrated into the legal interpretation grid, not the other way around. It is by reaffirming the judge’s mission, stating the law and deciding disputes, that a satisfactory balance between law and management science will be achieved.


* Gaëlle Deharo holds a doctorate in private law from the Université Paris I Panthéon-Sorbonne (2002) and she obtained her HDR (habilitation to supervise research) from the Université Versailles saint Quentin (2006). she continues her research in a rapidly expanding field: the articulation of law and management.



[1] Adam Nossiter, 35 Employees Committed Suicide. Will Their Bosses Go to Jail?, N.Y. Times (July 9 2019), https://www.nytimes.com/2019/07/09/world/europe/france-telecom-trial.html.

[2] Henry Samuel, France Telecom and Former CEO Stand Trial for ‘Harassment’ after 35 Employees Commit Suicide, The Telegraph (May 6, 2019), https://www.telegraph.co.uk/news/2019/05/06/france-telecom-former-ceo-stand-trial-company-wide-harassment; Angelique Chrisafis, Workplace Bullying Trial that Shocked France Draws to Close, The Guardian (July 8, 2019), https://www.theguardian.com/world/2019/jul/08/france-telecom-workplace-bullying-trial-draws-to-close; Nossiter, supra note 1.

[3] See J. Antonio Ariza-Montes et al., Workplace Bullying Among Managers: A Multifactorial Perspective and Understanding, 11 Int’l J. Envtl. Res. Pub. Health 2657 (2014).

[4] Moreover, many France Telecom workers were officials and could not be fired.

[5] Gaëlle Deharo & Sébastien Point, L’influence de la Recherché en Gestion sur la Qualification Juridique du Stress, 42 Revue Française de Gestion 39, 49 (2016).

[6] www.legifrance.gouv.fr.

[7] Id. (last visited on July 15, 2019).

[8] Code du Travail (Labor Code) art. L1152-1 (Fr.).

[9] See Ståle Einarsen, The Nature and Causes of Bullying at Work, 20 Int. J. of Manpower, 16 (1999); Izabel C. Martins Campos et al., Moral Harassment at Work Model and Inability, 41 Work 2060 (2012).

[10] See generally Emmanuel Abord de Chatillon, Le DRH de Demain Face aux Questions de Santé et Sécurité au Travail, 2 Management & Avenir 157 (2005).

[11] See generally M. Rajalakshmi & S. Gomathi, A Study on the Factors Influencing Workplace Bullying and its Impact on Employee Stress, 6 Mediterranean J. of Soc. Sci. 292 (2015); Mogens Agervold & Eva Gemzøe Mikkelsen, Relationships between Bullying, Psychosocial Work Environment and Individual Stress Reactions, 18 Work & Stress 336 (2004).

[12] See Juliana de Oliveira Barros et al., Précarisation du Travail et Souffrance Psychique: Le Cas du Service Pharmacie de l’Hôpital de l’Université de São Paulo, Brésil, 41 Travailler 143 (2019); Gemma Zontini, Réflexions sur la Dimension de l’Idéal dans les Troubles Psychiques Liés au Travail, 40 Travailler 65 (2018); Markus Hasel, A Question of Context: The Influence of Trust on Leadership Effectiveness during Crisis, 16 M@n@gement 264 (2013); Stewart R. Clegg & Ad van Iterson, The Effects of Liquefying Place, Time, and Organizational Boundaries on Employee Behavior: Lessons of Classical Sociology, 16 M@n@gement 621 (2013); Philip Bonifacio, An Overview of the Psychodynamic Approach, in The Psychodynamic Effect of Police Work: A Psychodynamic Approach 1-22 (1991).

[13] See Agervold & Mikkelsen, supra note 11; Nathan Bowling & Terry A. Beehr, Workplace Harassment From the Victim’s Perspective: A Theoretical Model and Meta-analysis, 91 J. Applied Psychol. 998 (2006); Lisa Bellinghausen & Nicolas G. Vaillant, Les Déterminants du Stress Professionnel Ressenti : Une Estimation par la Méthode des Equations d’Estimation, 195-196 Économie & Prévision 67 (2010).

[14] See Manuelle von Strachwitz, Prendre en Charge des Victimes de Harcèlement Moral, 23 Thérapie Familiale 251 (2002). 

[15] See Einarsen, supra note 9 at 17-20; Guy Notelaers et al., Measuring Exposure to Bullying at Work: The Validity and Advantages of the Latent Class Cluster Approach, 20 Work & Stress 289, 290 (2006); Marvin Claybourn, Relationships Between Moral Disengagement, Work Characteristics and Workplace Harassment, 100 J. Bus. Ethics 283, 283 (2011).

[16] See Marie-France Hirigoyen, Le Harcelement Moral: La Violence Perverse Au Quotidien (1998).

[17] Loi 2002-73 du 17 janvier 2002 de modernisation sociale [Law 2002-73 of January 17, 2002 on Social Modernization], https://www.legifrance.gouv.fr/affichTexte.do?cidTexte=JORFTEXT000000408905.

[18] Code pénal [C. pen] [Penal Code] art. 111-3 (Fr.). Official translation of the French Criminal Code available at https://www.legifrance.gouv.fr/Traductions/en-English/Legifrance-translations.

[19]  Code pénal [C. pen] [Penal Code] art. 111-4 (Fr.).

[20] See Cour de cassation [Cass.] [supreme court for judicial matters] crim., Dec. 8, 2009, Bull. crim., No. 09-84.117 (Fr.). All the references to French decisions that follow are available on the website www.legifrance.gouv.fr.

[21] Cour de cassation [Cass.] [supreme court for judicial matters] soc., June 26, 2019,  not published in bulletin, No. 17-20.723 (Fr.) (the judge decided that there was no moral harassment, based on the observation that there was no repetition, only isolated acts).

[22] The French Cour de cassation is the supreme court of France. It is in charge of the proper application of the law by the lower courts. If the law was not applied properly by a lower court, the Cour de cassation quashes the decision of the lower court and sends the case back to another court for judgment.

[23] Loïc Lerouge, La Constitution du Harcèlement Moral au Travail Indépendamment de l'Intention de son Auteur: Note Sous Cass. Soc. 10 Novembre 2009 No. 08-41.497, 28 Petites Affiches 18 (2010).

[24] Cour de cassation [Cass.] [supreme court for judicial matters] crim., Jan. 23, 2018, not published in bulletin, No. 16-87.709 (Fr.).

[25] Cour de cassation [Cass.] [supreme court for judicial matters] crim., Oct. 17, 2017, not published in bulletin, No. 16-86.075 (Fr.).

[26] Cour de cassation [Cass.] [supreme court for judicial matters] soc. (Fr.): Nov. 10, 2009, No. 07-45.321; Feb. 3, 2010, No. 08-44.107; Oct. 27, 2010, No. 09-42.488; Mar. 1, 2011, No. 09-69.616; Oct. 22, 2011, No. 13-18.862; Jan. 19, 2012, No. 10-18.059; June 20, 2012, No. 11-18.408; Jan. 17, 2013, No. 11-24.696; June 19, 2013, No. 12-18.850; June 19, 2013, No. 12-18.851; June 19, 2013, No. 12-18.854; June 19, 2013, No. 12-18.852; July 5, 2014, No. 13-11.038; May 21, 2014, No. 13-16.341. 

[27] See Brigitte Pereira, Du Harcèlement Moral au Harcèlement Managérial: Les Limites de l'Objectivation Légale, 39 Revue Française de Gestion 33 (2013).

[28] See id.

[29] See Fany Lalanne, Les Méthodes de Gestion d'un Supérieur Hiérarchique Peuvent Caractériser un Harcèlement Moral! La Cour de Cassation à l'Aune du Harcèlement "Managérial", La Lettre Juridique No. 374, Lexbase (Dec. 3, 2009); Christophe Radé, Harcèlement Managérial: La Cour de Cassation Plus Stricte?, Éd. Sociale No. 589, Lexbase Hebdo (Nov. 6, 2014); Christophe Radé,  Absence de Caractérisation d’un Harcèlement Moral Managérial à la Suite d’une Gestion Autoritaire et Inappropriée d’une Situation par un Supérieur Hiérarchique, Éd. Sociale No. 589, Lexbase Hebdo (Nov. 6 2014); Christophe Radé, Le Harcèlement Managérial de Nouveau Sanctionné, No. 434, Lexbase Hebdo (Mar. 31, 2011).

[30] Christophe Radé, Harcèlement Managérial: La Cour de Cassation Plus Stricte?, Éd. Sociale No. 589, Lexbase Hebdo (Nov. 6, 2014).

[31] Cour de cassation [Cass.] [supreme court for judicial matters] crim., May. 17, 2018, not published in bulletin, No. 17-83.889 (Fr.).

[32] Cour de cassation [Cass.] [supreme court for judicial matters] crim., June 19, 2018, not published in bulletin, No. 17-82.649 (Fr.).

[33] Cour de cassation [Cass.] [supreme court for judicial matters] crim., Nov. 13, 2018, not published in bulletin, No. 17-85.005 (Fr.); Cour de cassation [Cass.] [supreme court for judicial matters] crim., June 19, 2018, not published in bulletin, No. 17-86.737 (Fr.).

[34] Cour de cassation [Cass.] [supreme court for judicial matters] crim., Feb. 19, 2019, not published in bulletin, No. 18-83.268 (Fr.).

[35] Cour de cassation [Cass.] [supreme court for judicial matters] soc., Sept. 24, 2008, not published in bulletin, No. 06-45.772 (Fr.).

[36] Cour de cassation [Cass.] [supreme court for judicial matters] soc., Mar. 8, 2017, not published in bulletin, No. 15-24.406 (Fr.).

[37] Cour de cassation [Cass.] [supreme court for judicial matters] soc., May 10, 2012, not published in bulletin, No. 11-11.152 (Fr.).

[38] But see Cour de cassation [Cass.] [supreme court for judicial matters] soc., June 1, 2016, not published in bulletin, No. 14-28.870 (Fr.).

[39] See Dieter Zapf, Organisational, Work Group Related and Personal Causes of Mobbing/Bullying at Work, 20 Int’l J. Manpower 70 (1999).

[40] Sylvaine Perragin, Le Salaire de la Peine: Le Business de la Souffrance au Travail 81-89 (2019).

[41] Pascale Robert-Diard, Procès France Telecom: “On n’a Rien vu Venir. C’est Aussi Parce qu’on a Refusé d’Écouter”, Le Monde (June 21, 2019), https://www.lemonde.fr/societe/article/2019/06/21/proces-france-telecom-on-a-pousse-le-ballon-un-peu-trop-loin_5479867_3224.html.

[42]   What Business Life After the France Telecom Affair?, France Culture (July 5, 2019), https://www.franceculture.fr/emissions/linvite-des-matins/quelle-vie-en-entreprise-apres-laffaire-france-telecom.

[43] Claybourn, supra note 15. 

[44] See Annette Bool, L’accroissement des Plaintes pour Harcèlement Moral, est-il Inhérent au NPM?, XX(50) Revue Internationale de Psychosociologie et de Gestion des Comportements Organisationnels 155, 189 (2014).

[45] See Dieter Zapf et al., On the Relationship between Mobbing Factors, and Job Content, Social Work Environment and Health Outcomes, 5 European J. Work and Organizational Psychology 215 (1996).

[46] Elfi Baillien & Hans De Witte, Why is Organizational Change Related to Workplace Bullying? Role Conflict and Job Insecurity as Mediators, 30 Econ. & Indus. Democracy 348, 348-49 (2009); Gregor Bouville & Eric Campoy Une Approche Organisationnelle du Harcèlement Moral, 3 @GRH 53, 55 (2012).

[47] See Claire Bonafons et al., Specificity of the Links between Workplace Harassment and PTSD: Primary Results using Court Decisions, a Pilot Study in France, 82 Int’l Archives of Occupational & Envtl. Health 663 (2009).








Dispatch No. 20 - Spain - "Contradictory decisions on the employment status of platform workers in Spain"

by Alberto Barrio* 

January 2020



In May 2018, a judgement on the employment status of platform workers was issued for the first time in Spain. In the years since then, there have been at least seven other decisions by Spanish social courts of first instance deciding on this issue. In five judgements (concerning the platforms Take Eat Easy, Deliveroo and Glovo) it was concluded that the claimants were employees, while in three others (concerning the platform Glovo) it was maintained that they were self-employed workers. These judgements have alternated with decisions of the labour inspectorate, some of them stating that the riders of these online platforms were employees.

I. Legal framework on employment status in Spain

Two main legal statuses exist under Spanish law: employee (trabajador por cuenta ajena) and self-employed worker (trabajador por cuenta propia). Nevertheless, within the legal status of self-employment, a sub-variety exists, the economically dependent self-employed worker (trabajador autónomo económicamente dependiente), which grants certain rights that ‘regular’ self-employed workers lack. Claims on employment status classification are dealt with by the social courts of first instance, as well as by the labour inspectorate.

An employee under Spanish labour law is an individual who voluntarily performs remunerated work within the limits of the organisation and under the direction of another physical or juridical person.[1] This has been interpreted by the Supreme Court as to mean that an employment relationship requires dependency (indicated, inter alia, by the fact that the work is performed personally and in the times and locations established by the employer) and the performance of work for others’ behalf (indicated by, inter alia, the fact that the results of work is owned by the employer, and that it is the employer who fixes prices and salaries, and who selects clients).[2] Furthermore, there is a presumption of the existence of an employment contract when a person performs a service for another person (and within the sphere of management of that person) in exchange for remuneration.[3]

Self-employed workers, in turn, are defined as those individuals who carry out an economic or professional activity for profit, on a regular personal and direct basis, on their own account and outside the sphere of management and organization of another person, independently of whether or not they employ others.[4]

Economically dependent self-employed workers need to fulfil the abovementioned criteria, but they are also subjected to other requirements. In this regard, this form of work is defined as a self-employed person without employees who own their tools of production and who is financially dependent for at least 75 percent of their earnings on one client.[5] Economically dependent self-employed, unlike ‘regular’ self-employed, are entitled to annual leave, compensation if the main client does not fulfil the contract, coverage under the social court’s jurisdiction (instead of the civil court’s) and to join trade unions and be subjected to collective agreements (although not to participate in strikes), as well as easier access to the benefit for end of activity (similar to the unemployment benefit provided to employees). It is important to note that, however, this legal form is rarely used, with only 10,250 economically dependent self-employed workers registered as of September 2016.[6] This might be, at least in part, the result of burdensome regulation,[7] which requires, inter alia, that the contract is registered in the National Registry of Professional Associations of Self-Employed Workers (Registro Nacional de Asociaciones Profesionales de Trabajadores Autónomos).[8]

II. Judgements concerning the employment status of platform workers in Spain

The first judgment on the employment status of platform workers was handed down on the 29th May 2018 by the 11th social court of first instance of Barcelona, and concerned the dismissal of 20 riders performing work through the delivery platform Take Eat Easy, due to the end of activity of the company operating such platform, Take Eat Easy Spain S.L.[9] While this was the first judgement on the issue, it is important to note that it was preceded by several labour inspectorate decisions (mentioned below).

In the decision, the judge considered proven that the company required workers to use specific materials (backpacks with its logo and mobile phones with its app), as well as to set their working schedule four weeks in advance, and to respect certain standards of politeness. If the workers did not respect these requirements, their contractual relationship with the company could be terminated. Moreover, and unlike in later Spanish cases concerning the employment status of platform workers, the riders using Take Eat Easy were bound by an exclusivity clause with the company and received a minimum remuneration of 200 euros per month independently of the number of deliveries performed.[10] The case does not mention whether the riders were allowed to reject deliveries during an agreed shift of four hours. The judge concluded that the riders performed work under the control and within the sphere of organization of the company. As a result, it ruled that the claimants, most of them contracted as self-employed workers or through the intermediation of the cooperative Factoo (and two of them performing undeclared work) were instead to be considered employees of the company, and thus entitled to severance pay.

Almost simultaneously, on the 1st June 2018, the 6th social court of first instance of Valencia decided in a similar case with the same outcome, this time concerning one rider performing work through the platform Deliveroo.[11] This case addressed the situation of a rider whose contractual relationship was unilaterally terminated by the company operating Deliveroo due to an alleged lack of availability and an excessive rejection of deliveries within agreed shifts. The decision arguably explains in greater detail than the previous one how the conditions to which the rider was subjected pointed towards the existence of the two elements of the employment relationship (namely dependency and performing work for another party).  The existence of dependency of the rider from the platform was based on the fact that the rider was required to select his shifts (subjected to the company’s approval) with at least a week in advance, and would be sanctioned for any changes afterwards; the company (and not the rider) organised the professional activity through its app, which located at all time the rider (who was furthermore required to return to a certain ‘waiting’ place between shifts); that the rider received a minimun remuneration linked to his availability; and that the work was performed personally by the worker.[12] The fact that the rider performs work on behalf of the platform, in turn, was concluded based on the fact that the company operating the platform was the one which provided the main means of production (i.e. the platform itself and the app), established the price of the service (and provided bills on its name), required workers to use its uniform (as it told workers that they were “the image of the company in front of the client”[13]) and agreed on the delivery service with the restaurants. The judge ruled that the claimant must be considered an employee of the platform, and thus receive severance pay. This decision has since then become final, as the company operating the platform withdrew its appeal.[14]

The five judgements on the employment status of platform workers taken afterwards, from September 2018 until April 2019, concerned the company Glovoapp23 S.L., which operates the online platform Glovo. Like the abovementioned online platforms, this platform is used for the performance of delivery services, and the riders who work through it do so by signing a contract for the performance of services as economically dependent self-employed workers (with a working time of 40 hours per week), and by using the company’s app and their own vehicles (bicycle, motorbike or car). It is Glovoapp23 S.L., and not the rider, who operates the platform, who may make agreements with the stores from which products are delivered, who set the prices per delivery, and who collects the payment from and provide invoices to the recipient of the products (i.e. the client), as well as who pays the riders. Moreover, the riders need to select shifts in advance and are penalised (by having their access to delivery offers limited or, even, by the termination of their contract) if they are not available during their shift without permission. The company also provides a set of recommendations on how to acquire the products and how to relate with the clients. However, there are some differences in the conditions offered by Glovoapp23 S.L. compared to those offered by Take Eat Easy S.L. and Roofoods Spain S.L.

First, the company allows for the delivery of not only food, but also a wider range of services related to delivery (i.e. grocery delivery, collect and delivery of packages, and acquisition and delivery of products). As a result, the riders receive a company debit card, with which they may acquire products on the clients’ behalf, to whom they then deliver them. Second, the riders may choose which deliveries to perform among those offered to them or may, in turn, choose to be automatically assigned delivery orders. Third, the riders are not obliged to use materials with Glovo’s logo, but are provided with materials with this logo and are not allowed to use materials with other corporate logos (including their own). Fourth, the riders are rated by the clients, and this has an impact on the number of deliveries they are offered.

On the 3rd September 2018, the 39th social court of first instance of Madrid denied the claim of a rider that Glovoapp23 S.L. had wrongfully dismissed him due to temporary incapacity, based in part in the fact that he was considered to be correctly classified as an economically dependent self-employed worker.[15] The justification of the decision deviates in a few significant ways from the previous cases addressed by the 11th social court of Barcelona and the 6th social court of Valencia. In this regard, the judge considered that the rider was free to establish his own working time and availability, even though it did mention the potential reduction of deliveries offered to the rider if he did not respect the time slots that were previously approved by the platform.

It also considered relevant that the rider could decide the route for each delivery, and that the rider accepted the company’s proposal of changing the rider’s self-employed status to the one of economically dependent self-employed, a few months prior to the claim. In turn, the judge mentioned but did not consider key indicators of the existence of an employment relationship,  the facts that the platform fixed the prices for the service, requires its workers to use its app and provided its riders with debit cards. The decision did not mention the fact that Glovo riders use clothes and transport boxes with the company’s logo, as well as the fact that the riders’ working time is 40 hours per week. Moreover, the judge held that the means of transport (and not the online platform) was the main means of production, and thus that the means of production were provided by the rider.

On the 11th January 2019, in a very similar case, the 17th social court of first instance of Madrid also rejected a claim of wrongful dismissal due to temporary incapacity made by a Glovo rider, again partly based on the fact that the rider was considered to be an economically-dependent self-employed.[16] As in the previous case, the judge considered that it is the rider who provides the main means of production, who has a choice on whether or not to perform work (even though the court mentions that this may potentially produce less future deliveries for the rider), and who assumes the risk of the delivery (although it is not explained why it is considered so).

On the 11th February 2019, nevertheless, the 33rd social court of first instance of Madrid came to the opposite decision, considering that a Glovo rider was an employee.[17] The case considered that the unilateral termination of the rider’s contract within a context of a strike was a wrongful dismissal, as it violated the rider’s rights to freedom of expression and to strike. This judgement is arguably the first Spanish case in which the judge explicitly takes into account the new features of online platforms. In this regard, the judge starts his legal reasoning by analysing how the new information and communication technologies change fundamentally the way remuneration is set and evaluation is performed, as well as where and when work may be performed. The judge then notes that the company operating the platform has at its disposal a significant enough number of riders so that the freedom of the riders to choose their shifts does not have a significant impact on the business as a whole.

In contrast, the judge asserts, the rider could not perform such work without the online platform, which is considered as the main mean of production. Moreover, the judge states that the on-demand character of platform work, which is enabled by new IT tools, are vital to the rider’s capacity to set their working time as an indicator of employment status. Other elements that indicate the existence of an employment relationship which are mentioned, inter alia, are that the recipients of deliveries may not choose a particular rider and, most importantly, that it is Glovo who fixes the price of the service. Moreover, and although the court observed that the rider must assume the risk of not receiving payment for the delivery if the recipient is not satisfied, it also notes that it has not been proven that such a clause was in fact used. Furthermore, the decision refers to a set of international judgements (from the United States, the United Kingdom, France and Australia), as well as to a recommendation by the ILO and a Communication by the European Commission.

Some days after, on the 20th February 2019, the 1st social court of first instance of Gijón concluded that a rider of Glovo must be classified as an employee.[18] The judgment is mainly based on the facts that the platform is the main means of production, that it is the platform (and not the rider) that produces invoices for the service, and that the rider is in fact subordinated to the platform due to the rating system and the use of GPS technology to monitor the performance of the service.

On the 4th April 2019, the 1st social court of first instance of Madrid ruled that a Glovo rider was an employee, and considering the termination of his contract a wrongful dismissal. The decision, notably, reproduces verbatim many of the legal grounds of the decision by the 33th social court of first instance of Madrid.

Finally, on the 21st May 2019, a decision of the 24th social court of first instance of Barcelona rejected the claim of wrongful dismissal made by a Glovo rider, and stated that the rider was in fact correctly classified as an economically self-employed worker.[19] In this regard, the judge concluded that the rider of the platform has the capacity to manage their professional activity, which is inferred from the facts that they decide their own itinerary and working times (and that they may reject orders if they wish to do so), that they provide their own means of transport, and that they receive payment per delivery performed. Two elements, however, may differentiate this case from others. First, the case addresses the situation of a rider who stopped receiving delivery orders because the company considered that he was trying to deceive the app so it would believe that the distance between deliveries was greater than what it was in reality, an allegation which is not considered proven or disproven in the judgement. And, second, the judgement does not mention the parameters used to determine which delivery orders are offered to each rider, something which was covered in more detail by all other previous decision concerning the platform Glovo.

Besides the judicial decisions which have been analysed above, it is also worth noting the numerous labour inspectorate decisions taken across Spain on the employment status of platform workers. While this

contribution does not attempt to provide a comprehensive list of such interventions, some may be highlighted. In this regard, in November 2016 the labour inspectorate stated that riders using the platform Glovo in Barcelona were indeed economically dependent self-employed workers. [20] Nevertheless, later the labour inspectorates in Valencia,[21] Zaragoza[22] and, again, Barcelona,[23] considered that riders using this same platform must be considered employees of the company operating the platform. Furthermore, the labour inspectorates in the cities of Valencia,[24] Madrid[25] and Barcelona[26] have also stated that several of the riders performing work through the platform Deliveroo must be considered employees.

Following an appeal of a decision of the Madrid’s labour inspectorate, the Social Security Treasury (Tesorería General de la Seguridad Social) started a case before the social court of first instance of Madrid number 24 on this issue. The case, which directly involves over 500 riders using the platform Deliveroo, started its hearings on the 31st May 2019.[27] The outcome of this upcoming judgement may serve as a good indicator of the future of the discussion on the employment status of platform workers in Spain.

III. Conclusion

Some tentative conclusions might be drawn concerning both the main aspects highlighted in the abovementioned judgements and the general approach the courts have taken until now towards platform work.

Most cases revolve around the notion of whether the riders are free to set their own availability to work. Those decisions that found riders to be economically dependent self-employed workers considered that the riders were free to manage their working time and work availability, while those which reclassified them as employees stressed the underlying mechanisms restricting in practice their freedom (such as the use of sanctions when a rider rejects an order during a shift). In the same vein, decisions leaning towards self-employed status gave less relevance to the use by the riders of platforms’ uniforms and material than those which found riders to be employees.  Moreover, most decisions finding riders to be economically dependent self-employed workers seem to accept the company’s description of its main activity (i.e. the provision of IT services supporting the performance of deliveries), while the opposite decisions are often linked to considering restoration [SC1] or transport as the platforms’ main business.

Most judgements analysed do not highlight the arguably new nature of platform work. The decision of the 33th social court of first instance of Madrid is striking in this regard, for it not only acknowledges platform work’s new features, but it maintains that they are significantly different from the features of the traditional employment relationship, and that these differences justify a change in the way some indicators of the existence (or lack thereof) of an employment relationship are assessed (most importantly, the workers’ freedom to set their working time). It is now an enigma whether other judgements will follow a similar approach. By reproducing most of its legal arguments, the decision by the 1st social court of first instance of Madrid pointed in that direction. However, the decision of the 24th social court of first instance of Barcelona seemed to favour a more traditional application of the criteria to determine the existence of an employment relationship. Future judgements may provide more information on which path is taken.

*Alberto Barrio, PhD Candidate at the Labour Law and Social Policy Department of Tilburg University



[1] Real Decreto Legislativo 2/2015, de 23 de octubre, por el que se aprueba el texto refundido de la Ley del Estatuto de los Trabajadores, artículo 1(1).
[2] See, inter alia, Tribunal Supremo. Sala de lo Social, Sentencia núm. 902/2017, §3.
[3] Ibid, artículo 8(1).
[4] Ley 20/2007, de 11 de julio, del Estatuto del Trabajo Autónomo, artículo 1(1)
[5] Ibid., artículo 11.
[6] Rodríguez-Cabrero, A., ‘Spain’, ESPN Thematic Report on Access to social protection of people working as self-employed or on non-standard contracts, Brussels: European Commission, 2017, p. 7.
[7] Cherry, M., Aloisi, A., ‘Dependent Contractors in the Gig Economy: a Comparative Approach’, American University Law Review, vol. 66, p. 639.
[8] Ley 20/2007, de 11 de julio, del Estatuto del Trabajo Autónomo, artículo 11b.
[9] Juzgado de lo Social de Barcelona No. 11, sentencia de 29 de Mayo de 2018, ES:JSO:2018:2390.
[10] It is, however, unclear the impact of these two features on the final decision.
[11] Juzgado de lo Social de Valencia No. 6, sentencia de 1 Junio de 2018, ES:JSO:2018:1482.
[12] It is stated in the legal grounds of the case that, while, in theory, the rider could replace himself with another person, it was up to him to find the other person, and required the consent of the company to do so. Moreover, this clause was not used in practice by the claimant.
[13] Juzgado de lo Social de Valencia No. 6, sentencia de 1 Junio de 2018, ES:JSO:2018:1482, p. 7.
[14] EFE, ‘Deliveroo acepta la sentencia que consideró falso autónomo a un repartidor de Valencia’, El Pais, 6 November 2018.
[15] Juzgado de lo Social de Madrid No. 39, sentencia de 3 Septiembre de 2018, ES:JSO:2018:3042.
[16] Juzgado de lo Social de Madrid No. 17, sentencia de 11 de Enero de 2019, ES:JSO:2019:269. 
[17] Juzgado de lo Social de Madrid No. 33, sentencia de 11 de Febrero de 2019, Sentencia No. 53/2019.
[18] Juzgado de lo Social de Gijón No. 1, sentencia de 20 de Febrero de 2019, ES:JSO:2019:280.
[19] Juzgado de lo Social de Barcelona No. 4, sentencia de 4 de Abril de 2019, Sentencia No. 202/2019.
[20] In a report of the Labour Inspectorate of 4 November 2016, as noted in Juzgado de lo Social de Madrid No. 17, sentencia de 11 de Enero de 2019, ES:JSO:2019:269, p. 3.
[21] Zafra, I., ‘La Inspección de Trabajo en Valencia dice que 200 repartidores de Glovo son falsos autónomos’, El Pais, 27 December 2018.
[22] Ubieto, G., ‘Inspección de Trabajo dictamina en Barcelona que los 'riders' de Glovo son falsos autónomos’, El Periódico, 14 February 2019.
[23] Ibid.
[24] Inspección Provincial de Trabajo y Seguridad Social de Valencia, Acta de liquidación de cuotas a la seguridad social, desempleo, fondo de garantia salarial y formación professional por falta de afiliación o alta, dirigida a Roofoods Spain, SL., por el periodo 05/2016 al 09/2017, comunicada el 11 Diciembre 2017, 2017.
[25] Gómez, M. V., ‘La Inspección de Trabajo asesta otro golpe al modelo laboral de Deliveroo’, El Pais, 26 January 2018.
[26] Gómez, M. V., ‘La Inspección de Trabajo reclama a Deliveroo 1,3 millones de euros por sus falsos autónomos en Barcelona’, El Pais, 3 July 2018.
[27] Gómez, M.V., ‘Macrojuicio al modelo laboral de los ‘riders’ de Deliveroo’, El Pais, 31 May 2019.



Dispatch No. 19 - Germany - "EU law requires working time recording:
Consequences of the ECJ-ruling in the case 'CCOO' for member states and especially Germany"

by Dr. Thomas Klein* & Dominik Leist**

December 2019

In its decision of May 14, 2019, in case C-55/18, the European Court of Justice (ECJ) ruled that EU law requires member states to impose an obligation on employers to set up a system enabling the duration of daily working time to be measured.[1] This article first examines the main proceedings and the way in which the Court argues. In a second step, the concrete consequences of the judgement for the member states in general and the situation in Germany in particular are presented.

I. Underlying preliminary ruling procedure

In an increasing number of national court proceedings, the interpretation of national law depends on EU law requirements, the interpretation of which in turn is the sole responsibility of the ECJ. A reference for a preliminary ruling by a national court serves to enable national courts in such a situation to refer their question of interpretation of EU law in advance to the ECJ in order to consider its point of view in their decision.

The decision to be addressed is based on such a reference for a preliminary ruling from the Audiencia Nacional (National Court of Justice), Spain. It was preceded by an action brought by the CCOO, an association of workers, seeking a declaration that Deutsche Bank as employer was required under Spanish law to set up a system to record the actual daily working hours of all their workers. The CCOO argued that an obligation to establish such a system arises from the interpretation of Spanish law in the light of Article 31(2) CFREU (Charter of Fundamental Rights of the European Union ) and the European Working-Time-Directive (Directive 2003/88/EC).[2] However, it had not yet been clarified whether the EU law requires such an obligation.

The Audiencia Nacional therefore referred several similar questions to the ECJ, which the ECJ summarized in the question whether Art. 3, 5, 6, 16, 22 of Directive 2003/88/EC in conjunction with Articles 4(1), 11(3) and 16(3) of Directive 89/391/EEC and Article 31(2) of the CFREU must be interpreted as precluding the legislation of a member state which, according to its interpretation by the national courts, does not oblige employers to set up a system measuring the daily working time worked by each worker.[3][4]

II. Decision and reasoning of the ECJ

The ECJ ruled that Art. 31 (2) CFREU and the aforementioned directives preclude a national regulatory regime which does not oblige the employer to record working time. Neither Art. 31 (2) CFREU nor the directive explicitly provide for a recording obligation. However, the ECJ argues that such an obligation is necessary in order to guarantee the practical effectiveness of expressly standardized rights such as the weekly and daily minimum rest periods and the weekly maximum working time.[5] In particular, the ECJ emphasizes that without such a system the worker can hardly effectively prove his actual hours worked.[6] In this context, the ECJ again emphasizes that the directive standards to be examined concretize the fundamental right enshrined in Art. 31 (2) CFREU and must therefore be interpreted in its light.[7] The ECJ also establishes concrete requirements for the necessary time recording. The necessary practical effectiveness of the envisaged rights is only guaranteed if employers are obliged to set up an "objective, reliable, and accessible" system for recording working time.

III. Consequences of the ruling for the Member States

For the member states, the first result is that the implementation obligations with regard to the directive requirements also include the obligation of the employers to introduce a working time recording system. Since the transposition deadline has already expired, the member states must immediately establish an obligation to record working time, unless such an obligation already exists explicitly or can at least be assumed by interpreting national law.

The requirements of the ECJ expressly demanding an "objective, reliable, and accessible" system must be adhered to. The exact meaning of these features is currently the subject of lively discussions.[8] It is not least questionable whether the feature "objective" refers to the recording itself or to the usability of the recording result.[9] In designing such a system, however, the ECJ expressly leaves the Member States a margin of discretion.[10] However, this refers only to the "how", not to the "whether" of an obligation to record working time. The base limit here is the suitability for practical effectiveness of directive's rights.

Another interesting question at present is whether the judgement also imposes an obligation on private employers to record working time even before the court’s requirements are implemented by the member states. This could, on the one hand, be the case if the provisions of the Working Time Directive could be applied directly between private individuals. EU-Directives basically only serve as targets that have to be implemented by the Member States within a given deadline. The ECJ has indeed provided for exceptions to this principle in previous decisions in cases where the transposition deadline has passed and “unconditional and sufficiently precise rules” in the Directive exist at the same time.[11] These exceptions, however, relate only to an effect on public authorities. [12] In contrast, a direct effect of EU-Directives between private individuals cannot be considered according to the settled case law of the ECJ.[13]

The alternative possibility for an obligation on employers to record working time even before the implementation of the judicial provisions in national law would be to regard the fundamental right of Article 31 (2) CFREU as directly applicable in the horizontal relationship (between private individuals). This interpretation has already been affirmed in another decision regarding the holiday periods also granted in Art. 31 (2) CFREU.[14] However, the ECJ does not comment on the direct applicability of the fundamental right in a horizontal relationship. This is presumably due to the narrowly framed question referred for the ruling and does not imply that the judgment is to be understood as a rejection of this option. Quite to the contrary, the conditions laid down in the case "Max Planck Gesellschaft zur Förderung der Wissenschaften" for a direct horizontal effect, as already emphasized by Advocate General Pitruzzella , speaks clearly in favor of the affirmation of a direct horizontal effect also in this case, i.e. with regard for working time law.[15][16]

IV. Need for implementation in Germany

German law provides for time recording obligations only selectively for certain professions specifically susceptible to abuse. In comparison, a general recording obligation - similar to Spanish law - is only provided for overtime, but not for regular working hours.

The question therefore arises as to whether German law can be interpreted in conformity with EU law in the sense of a general obligation to record working time. As the statute to be interpreted accordingly might be considered the ArbZG (German Working Time Act), which is part of the public law on occupational health and safety. Violations of the provisions are subject to fines in accordance with §§ 22, 23 ArbZG. Because of the punitive nature of those fines an analogous application of the recording obligation for overtime is opposed to the constitutional prohibition of analogy for criminal law and the imposition of fines (Article 103 section 2 of the Basic Law).[17]

German civil law could, however, be interpreted in conformity with EU law without methodological reservations. For example, one could regard the obligation under EU law as a contractual secondary obligation of the employer within the meaning of § 241 (2) BGB (Civil Code) or as part of the obligation under § 618 (1) BGB (Civil Code).[18] In both cases, however, the enforcement obligation would lie with the employee who would have to sue their employers for setting up a recording system. This cannot satisfy the requirements of the ECJ, which in its judgment repeatedly emphasizes the weak position of the employee vis-à-vis his employer, as a need for providing an effective regulation. Alternatively, rights of initiative and enforcement possibilities of the work council for the introduction of a recording system might also be an option.[19] Thus the right of codetermination from § 87 (1) No. 2 BetrVG (Works Constitution Act) refers also to the recording and control of the working time or the renunciation of such and contains a right of initiative of the work council.[20][21] Although these possibilities can be helpful in some cases, they do not provide the actual effectiveness necessary as in many enterprises, no work council exists.

Altogether, therefore, German law cannot be interpreted in a way that meets the requirements of the ECJ for the recording of working time, which effectively means that German law does not currently comply with EU law. A legislative intervention for introducing a time recording system in compliance with the requirements of the European Court of Justice is therefore an urgent necessity.

*Dr. Thomas Klein, Research Associate at the Institute for Labour Law and Industrial Relations in the European Union (IAAEU), Trier, Germany
Contact: klein@uni-trier.de
**Dominik Leist, Research Associate at the chair of Professor Monika Schlachter, Law Faculty, Trier University, Germany
Contact: leist@uni-trier.de


[1] Case C-55/18 Judgment, Federación de Servicios de Comisiones Obreras (CCOO) v. Deutsche Bank SAE (May 14, 2019), http://curia.europa.eu/juris/documents.jsf?num=C-55/18.

[2] Charter of Fundamental Rights of the European Union, art. 31(2), 2012 O. J. (C 326) 391, 401.

[3] Council Directive 2003/88, arts. 3, 5, 6, 16, 22, 2003 O.J. (L 299) 9, 11-15 (EC).

[4] Council Directive 2003/88, arts. 4(1), 11(3), 16(3), 1989 O.J. (L 183) 1, 3-7 (EC).

[5] Case C-55/18 Judgment, CCOO v. Deutsche Bank SAE, ¶¶ 46-60 (May 14, 2019), http://curia.europa.eu/juris/documents.jsf?num=C-55/18.

[6] Id. ¶ 54.

[7] Id. ¶ 31 (citing Case C-112/13 Judgement, A v. B and Others, ¶ 51 (Sept. 11, 2014), http://curia.europa.eu/juris/documents.jsf?num=C-112/1; Joined Cases C-569/16 and C-570/16 Judgement,  Stadt Wuppertal v. Maria Elisabeth Bauer and Volker Willmeroth v. Martina Broßonn ¶ 85 (Nov. 6, 2019), http://curia.europa.eu/juris/documents.jsf?num=C-569/16.

[8] See Thomas Klein & Dominik Leist, Die Unionsrechtliche Pflicht zur Arbeitszeiterfassung, Zeitschrift für Europäisches Sozial- und Arbeitsrecht, 2019, at 367.

[9] Id.; Tortsten von Roetteken, Pflicht des Arbeitgebers zur Einrichtung eines Systems zur Erfassung der Täglichen Effektiven Arbeitszeit, 23 Juris-Praxisreport Arbeitsrecht 1, at D (2019).

[10] Case C-55/18 Judgment, CCOO v. Deutsche Bank SAE, ¶ 63 (May 14, 2019), http://curia.europa.eu/juris/documents.jsf?num=C-55/18.

[11] Instead of all. Case-8/81 Judgement, Ursula Becker v. Finanzamt Münster-Innenstadt, ¶ 25 (Jan. 19, 1982), http://curia.europa.eu/juris/documents.jsf?num=C-8/81.

[12] Dominik Leist, Unionsrecht gebietet Pflicht zur Systematischen Arbeitszeiterfassung, 22 Juris-Praxisreport Arbeitsrecht 1, at D.II (2019) (rejecting the possibility of an immediate directive effect of the Working-Time-Directive also for state employers).

[13] Settled case law since Case C-152/84 Judgment, M. H. Marshall v. Southampton and South-West Hampshire Area Health Authority (Feb. 2, 1986), http://curia.europa.eu/juris/documents.jsf?num=C-152/84.

[14] Case C-684/16 Judgment, Max-Planck-Gesellschaft zur Förderung der Wissenschaften e.V. v Tetsuji Shimizu, ¶ 81 (Nov. 11, 2018), http://curia.europa.eu/juris/documents.jsf?num=C-684/16.

[15] Case C-55/18 Opinion, CCOO v. Deutsche Bank SAE, ¶ 94 (Jan. 31, 2019), http://curia.europa.eu/juris/documents.jsf?num=C-55/18.

[16] Johannes Heuschmid, Neujustierung des Arbeitszeitrechts und des Systems der Arbeitszeiterfassung durch den EuGH, Neue Juristische Wochenschrift, 2019, at 1854 Is 1854 a page number, and is it the pincite? – ask author; Thomas Klein & Dominik Leist, supra note 8, at 370; Roetteken, supra note 9, at C; Daniel Ulber, Arbeitszeiterfassung als Pflicht des Arbeitgebers – Die Folgen der Entscheidung des EuGH in der Rechtssache CCOO, Neue Zeitschrift Für Arbeitsrecht, 2019, at 680.

[17] Leist, supra note 12, at D.I.; Ulber, supra note 16, at 680 (not exluding an interpretation consistent with EU law).

[18] Klein & Leist, supra note 8, at 371.

[19] Id.

[20] R. Richardi, in Betriebsverfassungsgesetz: BetrVG § 87 para. 297 (R. Richardi ed., 16th ed. 2018); Thomas Klebe, in BetrVG: Betriebsverfassungsgesetz § 87 para. 100 (Wolfgang Däubler et al. eds., 16th ed. 2018). I would think page numbers would be better here – ask author.

[21] Richardi, supra note 20, at para. 65.




Dispatch No. 18 – Australia – "The development of labor hire regulation in Australia: Promoting the integrity of the labor hire industry and protecting workers engaged in labor hire arrangements through a licensing system," by Maria Azzurra Tranfaglia*

 August 7, 2019

This dispatch will start by exploring the rationale underpinning the recent introduction of labour hire legislation in some Australian states. It will then offer a general overview of these regulatory developments, namely: the Victorian Labour Hire Licensing Act 2018 (Vic), which has come fully into effect on 29 April 2019; the Queensland Labour Hire Licensing Act 2017 (Qld), in operation since June 2018; and the South Australian Labour Hire Licensing Act 2017 (SA). Final reflections will highlight the challenges faced by the legislators in defining the scope of application of the labor hire licensing regimes.

From an Australian perspective, the provision of a licensing regime for certain types of triangular arrangements represents an important change in the regulatory landscape aimed at addressing the lack of effective protections for workers engaged through unscrupulous and/or insolvent labor suppliers. At the same time, the Australian states’ legislative experience is a powerful illustration of the need to regulate labor intermediaries beyond pure labor hire or “agency work,” as it is more commonly referred to in Europe. It therefore constitutes a regulatory laboratory to watch closely when looking for solutions to better protect workers and the economy by the abuse of broader triangular business structures and fissuring mechanisms.


I. Labor hire arrangements: definitions, general trends, and the call for regulation

In the absence of a specific definition of labor hire under Australian labor law, the term has often been used loosely to encompass a multitude of triangular arrangements. Many have drawn the parallel with agency work as it is understood in Europe, i.e., where the formal employer is the employment agency and the worker is directed by the user-company/host.[1] At the same time, in in certain sectors (e.g. cleaning), labor hire has often been associated with labor contractors, i.e., third parties that may or may not maintain control over their workforce.[2]

Labor hire arrangements have generally been referred to as “triangular relationships” in which a labor hire agency supplies the labor of a worker to a third party (host) in exchange for a fee.[3] In practice in Australia, a labor hire worker is engaged either as an independent contractor or, more often, as a casual employee by an agency that is then hired under a commercial agreement to another business that coordinates or controls that worker’s activity. Whilst the engagement of labor hire independent contractors has long been a controversial, though accepted practice, these sort of arrangements—also referred to as “Odco” style arrangements from the name of the business that started using them—have lately been deemed fraudulent unless highly skilled workers are involved.[4] On the other hand, in cases involving labor hire employees, according to the common law, the exercise of control by the host is an inherent characteristic of labor hire arrangements.[5] Therefore courts have generally found the labor hire agency to be the true employer, unless the latter was found to act as a mere screen between the host and the workers.[6]

So far, Australian labor hire agencies have ranged from the large well-known multinational firms, such as Adecco and Manpower, and some substantial local firms, such as Skilled Engineering, to many small operators.[7] As recently summarized in the Report of the Victorian Inquiry into Labour Hire and Insecure Work, the reality of labor hire appears polarized:

At one end of the spectrum are labour hire agencies and arrangements which are highly transparent and compliant with workplace laws, awards and other industrial instruments, health and safety legislation and other applicable legal requirements. At the other end of the spectrum are ‘invisible’ labour hire agencies and arrangements, operating almost entirely outside the existing regulatory framework. However, the boundaries between them are not clear. There is a range of agencies and arrangements falling between the two extremes. . . . Sometimes, the reputable and rogue operators interact. . . .[8]

Australia has never signed the ILO Convention No. 181/1997 on private employment agencies[9] and up until recently many scholars have noted a general lack of systematic protective regulation for intermediated workers, resulting in a highly porous and opaque system.[10] Despite the acknowledgement of the importance and the legitimacy of labor hire arrangements for the Australian economy due to the need for flexibility in modern workplaces, concerns have often been raised in the past two decades in relation to the effective protection of the workers involved, especially in light of their predominantly casual status.[11] Clauses ensuring equal treatment of on-hired workers have become a feature of Modern Awards, a set of 122 federal instruments developed between 2008 and 2010 that establish industry or occupational minimum employment standards.[12] However, labor hire workers remain not legally entitled to the same rights established for user companies’ employees under the relevant enterprise agreements, unless unions manage to negotiate parity clauses.[13] Moreover, there are generally no legal thresholds in relation to the number of labor hire workers that a company can host. Most importantly, in case of insolvent fly-by-night operators, labor hire workers cannot automatically claim damages against the host. User companies can be sanctioned as accessories if their intentional involvement in the contravention can be proved.[14] Nevertheless, under Australian law the responsibility for underpayments in case of triangular work arrangements generally lies with one entity: the direct employer. And this approach has begun to appear increasingly out of sync with the current fragmented world of work relationships, where workers are often dependent on and/or controlled by more than one entity.[15]

A series of inquiries both at State and at Federal have been promoted in the past to assess the adequacy of the law in protecting labor hire workers from unscrupulous practices aimed at undercutting employment conditions.[16] They triggered much research interest but no definitive regulatory solution.[17] The introduction of a licensing system for labor hire providers has often been placed on the agenda, mainly by the Union movement, but also by agencies interested in improving the reputation of the industry.[18] However, such proposals have never received much support from governments in the past and the view that they would impose unnecessary burdens and be unlikely to eradicate fly-by-night operators prevailed.[19]


II. The State Labor hire licensing Acts and their scope of application

Between 2014 – 2016, a series of media revelations of exploitative practices associated with triangular arrangements such as franchising and labor hire in certain sectors (including but not limited to cleaning, horticulture, and food processing) have forcefully brought the issue to the public attention and led to a revival of calls for regulatory reforms both at State and at Federal level.[20] The Fair Work Ombudsman started conducting a series of strategic enforcement activities aimed at encouraging compliance with workplace rights, including investigations of high profile businesses engaging in labor hire arrangements with unscrupulous providers.[21] The federal regulator’s attempts to keep pace with the evolving landscape of triangular business structures[22] built momentum for the introduction at federal level of an expanded form of accessorial liability entailing higher sanctions applicable to franchisors and holding companies.[23] At the same time, inquiries into the labor hire industry have been conducted in a number of States, including Victoria, South Australia, and Queensland, and have recommended the introduction of a licensing scheme to improve the reliability of labor market intermediaries.[24] These investigations were all established between 2015 and 2016 to measure the extent, the nature and the consequences of labor hire and to then recommend the best regulatory response. They have raised important questions in relation to whether and to what extent the major players overseeing these business structures should be deemed liable as entities that ultimately dictate the working conditions and profit from the labor of the workers at the base.[25] The support for a licensing scheme at state level was significant and generally underpinned by a narrative that attempted to reconcile the need of addressing the protection gap for workers with the importance of restoring trust in the labor hire industry for the benefit of businesses and the overall economy.

The Victorian independent Inquiry recommended the introduction of a licensing scheme modeled on the U.K. Gangmasters (Licensing) Act in the sectors characterized by the most egregious breaches of workplace rights: horticulture, meat processing, and cleaning.[26] The Victorian Government accepted many of the Inquiry’s recommendations. However, it opted for a universal licensing scheme, similar to those initially enacted under the South Australian Labour Hire Licensing Act 2017 (SA), and the fully operational Queensland Labour Hire Licensing Act 2017 (Qld). Though this type of licensing approach has been opposed and criticized by representatives of employment agencies’ peak body for unfairly capturing an unnecessarily imposing high licensing fees on white-collar businesses, the independent chair of the Victorian Inquiry has welcomed its adoption.[27] As of 29 April 2019 the Victorian Act has effectively come into force and labor hire providers have six months from that date to apply for a license to operate legally. In regard to the South Australian regime minted in 2017, following the election of the new Liberal government and a re-examination due to concerns around the effectiveness of the scheme, its operation was postponed. However, the attempt to repeal the Act failed and the scheme has been revived, with a new deadline being set on 31 August 2019 for businesses to apply for the relevant license.

The general structure of the labor hire licensing schemes, though with some peculiarities for each state,[28] can be summarized as follows:

  • providers of labor-hire services are required to hold a license, and businesses wanting to use labor-hire staff are required to only use licensed providers;
  • to obtain a license, providers are required to pass “a fit and proper person test,” and show, among other things, to be a financially viable business and to have a record of compliance with workplace and labor-hire laws;
  • licensed providers are subject to reporting obligations;
  • an authority is responsible for monitoring and investigating compliance;[29]
  • operators that do not comply with or attempt to elude the scheme are liable for civil and criminal penalties.

In other words, the regulatory regimes operate to deter businesses from engaging unreliable providers and to eradicate unscrupulous practices by establishing entry requirements in an industry where the commodity being sold is people’s labor. To this end, they impose some level of responsibility on the client/user companies by requiring them to only engage licensed providers and establishing sanctions for failure to comply with the scheme.[30]

The legislators have been grappling with the difficulties of defining what labor hire services providers are, while trying to capture the multifaceted reality of triangular employment arrangements and the complexities of the contracting structures in sectors that are most at risk of noncompliance. Both the South Australian and the Victorian Act identify a “provider of labour hire services” as any person that in the course of carrying on a business supplies to another person a worker to do work “in and as part of the business . . . of the other person.”[31] This is the case regardless of who controls the work (host or labor hire services provider) and regardless of whether the workers are supplied directly or through one or more intermediaries.[32] Moreover, the Victorian Act extends the application of the regime also to operators that merely perform the recruitment and placement function, if they are providing accommodation for the period during which the individuals perform the work, as well as to firms that offer contractor management services.[33] The general definition appears even broader in Queensland, where “a person (provider) provides labour hire services if, in the course of carrying on a business, the person supplies another person (a worker) to do work.”[34] The statutes above also contain a very extensive notion of “worker” (not just employees), with the Victorian Act expressly including also apprentices and individuals under training contracts.[35]

The letter of the law might prima facie suggest that a much broader range of triangular arrangements beyond pure labor hire (e.g. subcontracting, secondment arrangements, as well as other work models typical of the so-called gig economy) are to be covered by the licensing requirements. However, the risk of overreach is partly counterweighed by the current interpretation of what constitute labor hire as opposed to genuine subcontracting.[36] Furthermore, a series of exceptions have been (and will potentially continue to be) introduced by relevant regulations aimed at excluding class of labor hire providers, services, or workers. In Queensland, for example, where the licensing scheme is fully operational since June 2018, the regulations have excluded, amongst others, the application of the regime for the provision of workers earning more than the high-income threshold (currently $145,400) and lawyers seconded to a client.[37]

In South Australia the Government Gazette dated 6 June 2019 has carved out a series of exemptions including “where the business provides workers:

  • but providing labour hire services is not a core function of the business
  • to another business within the same group of companies – e.g. parent company supplying workers to a subsidiary company
  • to work in another business where both businesses are part of the same franchise
  • to work in another business that are not part of a franchise but are collectively operating using the same banner, branding or trading name.”[38]

Under the Victorian regime, instead, the Regulations have excluded amongst others i) secondment arrangements, ii) arrangements in which a worker performs work for an entity or group of entities that carry on business collectively as one recognizable business and iii) workplace learning and vocational placements.[39] On the other hand, the Victorian Regulations have provided a series of deeming provisions that essentially extend the application of the labor hire licensing regime to any type of triangular arrangement concluded in the cleaning, horticulture, and meat industries.[40] This addition was arguably necessary to prevent that arrangements in industries found to be the most problematic by the Victorian Inquiry, might be disguised as genuine subcontracting arrangements to elude the application of the new legislation.



In summary, though arguably not depicting the traditional understanding of labor hire/agency work, the Acts have tried to capture the multifaceted reality of triangular employment arrangements and the complexities of the contracting structures in sectors that are most at risk of noncompliance with workplace rights. To use a mythological metaphor, in tackling the notion of labor hire State legislators were, like Hercules, faced with the proverbial hydra.[41] In order to avoid the risk of multiple heads re-growing outside the scope of the legislation due to a narrow predetermined definition, they have taken a broad initial stance to allow a case-by-case heuristic reduction and/or ad hoc shaping of the scope.

From a regulatory perspective, Australian state-based licensing schemes represent an innovative ex ante measure aimed at preventing, rather than remedying ex post, the negative impact of unscrupulous practices associated with certain triangular work arrangements. They differ from the sector specific U.K. regulation of gangmasters. At the same time, they are unlike other continental European countries’ regulatory models, where pure agency work authorization systems are coupled with a(n) (often difficult to enforce) strict joint-liability regime applicable to broader triangular arrangements, including subcontracting.[42] While the deterrent function of the penalties—together with the enforcement efforts of the respective licensing authorities—should ultimately benefit workers by increasing the likelihood of them dealing with reliable employers, state Labour Hire Licensing Acts cannot, due to jurisdictional limitations, create individual rights vis-a-vis user companies in case of underpayments. To this end, a more comprehensive framework at federal level would be needed. Some commentators have hypothesized the extension of the new liability regime introduced for franchisors and holding companies also to host companies in labor hire arrangements, and more specifically to lead companies in supply chains (including cleaning supply chains).[43] In the meanwhile, unions have tried to address the accountability issue by way of collective agreements with lead companies aimed at ensuring that labor hire providers in supply chains operate in compliance with employment standards.[44] In terms of policy proposals, the Labor party had placed equal treatment of labor hire workers and a national labor hire licensing system on the federal agenda.[45] Also the current Coalition government has recently endorsed a regulation of labor hire at national level, though in the form of a “light touch” national registration scheme limited to high risk sectors as recommended by the Report of the “Migrant Workers Taskforce.”[46]

* Maria Azzurra Tranfaglia is a PhD candidate at the University of Melbourne and acknowledges the support of an Australian Government Research Training Program. Azzurra is currently a Victorian public sector employee. However, the views expressed in this dispatch are personal and related to her PhD research and do not represent those of her current employer.


  1. [1] “What is termed ‘labour hire’ in the Australian context is referred to as ‘agency work’ in the United Kingdom and much of continental Europe.” See Anthony Forsyth, The Victorian Inquiry into Labour Hire and Insecure Work: Addressing Worker Exploitation in Complex Business Structures, 6 E-J. Intl. & Comp. Lab. Stud. 1, 6, (September-October 2017).
  2. [2] See, for example, the broad definition offered by Andrew Stewart et al., Creighton and Stewart’s Labour Law (6th ed. 2016).
  3. [3] Richard Johnstone et al., Beyond Employment: The Legal Regulation of Work Relationships 60 (1d ed. 2012).
  4. [4] See for example Fair Work Ombudsman v Quest South Perth Holdings Pty Ltd (2015) 228 FCR 346 (Federal Court of Australia); Fair Work Ombudsman v Contracting Solutions Australia Pty Ltd (2013) FCA 7 (Federal Court of Australia).
  5. [5] See, for example, FP Group Pty Ltd v Tooheys Pty Ltd (2013] FWCFB 9605, 28 (Fair Work Commission of Australia). See also, Fair Work Ombudsman v Ramsey Food Processing Pty Ltd (2011] FCA 1176, 60-61 (Federal Court of Australia):
  6. [6] FP Group Pty Ltd v Tooheys Pty Ltd (2013) FWCFB 9605 (Federal Work Commission of Australia). In this full Bench decision, the Fair Work Commission made clear that labor hire arrangements are not illegitimate per se, even when their objective is to avoid liability for employment-related obligations. It stated that the interposition of the hiring agency between its clients and the workers—which is the typical scheme of labor hire—does not automatically result in an employment relationship between the worker and the user-company. That a crucial consideration when assessing whether or not the labor supplier is the real employer of the worker it supplies is the commercial practicality of its business and the authenticity of the underlying arrangements.
  7. [7] John Burgess & Julia Connell, International Aspects of Temporary Agency Employment, in International Perspectives on Temporary Agency Work 1 (John Burgess & Julia Cornell eds. 2004).
  8. [8] Anthony Forsyth, Victorian Inquiry into the Labour Hire Industry and Insecure Work: Final Report, Department of Economic Development, Jobs, Transport and Resources: Melbourne 56 (August 2016).
  9. [9] See Elsa Underhill. The Role of Employment Agencies in Structuring and Regulating Labour Markets, in Labour Law and Labour Market Regulation (Christopher Arup et al. eds. 2006) at 304.
  10. [10] Iain Campbell et al., Temporary Agency Work in Australia: A Basic Profile and a Few Questions, in International Perspectives on Temporary Work and Workers (John Burgess & Julia Connell eds. 2004); For an historical perspective on employment agencies regulation in Australia See Anthony O’Donnell & Richard Mitchell, The Regulation of Public and Private Employment Agencies in Australia: An Historical Perspective, 23 Comp. Lab. L. & Pol'y J. 7 (2001).
  11. [11] Id. See also Pauline Thai, Unfair dismissal protection for labour hire workers? Implementing the doctrine of joint employment in Australia, 25 Au. J. Lab. L. 158 (2012).
  12. [12] Richard Johnstone & Andrew Stewart, Swimming Against the Tide? Australian Labour Regulation and the Fissured Workplace, 37 Comp. Lab. L. & Pol'y J. 55, 60 (2015).
  13. [13] “Parity,” “site rate,” and “jump up clauses,” which extend to labor hire workers the same conditions applicable to direct employees, are a common feature in unionized sectors. However, as labor hire employees are not directly covered by these agreements, such clauses can only be enforced by a union that is covered by the agreement.
  14. [14] Fair Work Act 2009 (Cth) s. 550.
  15. [15] Tess Hardy, Who Should Be Held Liable for Workplace Contraventions and On What Basis?, 29 Au. J. Lab. L. 78 (2016); Maria Azzurra Tranfaglia, Tackling Underpayment in Cleaning Supply Chains, Pursuit The University of Melbourne (July 12, 2019) available at https://pursuit.unimelb.edu.au/articles/tackling-underpayment-in-cleaning-supply-chains.
  16. [16] See Parliament of New South Wales, NSW Labour Hire Task Force Report (2001); Parliament of Australia, Inquiry into Independent Contractors and Labour Hire (2005). See also Brian Howe Et Al., Lives on Hold: Unlocking the Potential of Australia’s Workforce (April 2012). The 2012 Inquiry recommended that industrial regulations be strengthened to provide a universal set of protections to all Australian workers such as “Reforms to better capture indirect employment arrangements like labour hire and dependent contracting.”
  17. [17] Andrew Stewart, Redefining Employment? Meeting the Challenge of Contract and Agency Labour, 15 Au. J. Lab. L. 1 – 42 (2002); Steve O'Neill, Labour Hire: Issues and Responses, Parliament of Australia, Economics, Commercial and Industrial Relations Group, March 8, 2004, available at, http://www.aph.gov.au/About_Parliament/Parliamentary_Departments/Parliamentary_Library/pubs/rp/rp0304/04rp09; Trina Malone, Vulnerability in the Fair-work Place: Why Unfair Dismissal Laws Fail to Adequately Protect Labour-hire Employees in Australia, Centre for Employment and Labour Relations Law, The University of Melbourne 1, 24 (May 2011); Underhill Elsa, Report – Labour Hire Employment and Independent Contracting in Australia: two Enquiries, How Much Change? 19 Au. J. Lab. L. 306-14 (2006). Thai Pauline, Unfair Dismissal Protection for Labour Hire Workers? Implementing the Doctrine of Joint Employment in Australia 25 Au. J. Lab. L.\ 158 (2012).
  18. [18] Underhill, supra note 18, at 307.
  19. [19] See, for example, Skilled, Submission to the Senate Inquiry into Independent Contractors and Labour Hire, Parliament House, 2005, Canberra and Recruitment and Consulting Services Association (RCSA), Submission to the Senate Inquiry into Independent Contractors and Labour Hire, Parliament House, 2005, Canberra, cited by Mitlacher, Lars W. Mitlache & John Burgess, Temporary Agency Work in Germany and Australia: Contrasting Regulatory regimes and Policy Challenges 23 Int'l J. Comp. Lab. L. & Indus. Rel. 401, 427(2007). In particular, the key industry Group (RCSA) has repeatedly claimed that self-regulation would be the best option, often suggesting that a higher level of regulation would impose unnecessary burdens, be unlikely to eradicate fly-by-night operators, and ultimately reduce the number of jobs.
  20. [20] See ABC Four Corners, 7 -Eleven: The Price of Convenience (August 30, 2015). Tess Hardy, Franchises: All Care and No Responsibility, The Conversation (October 29, 2015). See also Pat McGarth, Myer Cleaners Accuse Retailer of Underpayment, Denying Entitlements with 'Sham Contracting' Practice, ABC News (October 22, 2015); Maria Azzurra Tranfaglia, Law Allows Myer to Outsource Responsibility for Labour Hire Workers, The Conversation (October 28, 2015); Maria Azzurra Tranfaglia, Australian Dream a Nightmare for Many Labour Hire Employees, The Conversation (February 18, 2015).
  21. [21] See Fair Work Ombudsman v Al Hilfi (2016) FCA 193 (Federal Court of Australia); Natalie James, A Report on the Fair Work Ombudsman’s Inquiry into the Labour Procurement Arrangements of the Baiada Group in New South Wales, Fair Work Ombudsman, available at https://www.fairwork.gov.au/about-us/news-and-media-releases/2015-media-releases/june-2015/20150618-baiada-group-statement-of-findings (June 2015). See also, Tess Hardy, Reconsidering the Notion of ‘Employer’ in the Era of the Fissured Workplace: Traversing the Legislative Landscape in Australia, in 95 Bulletin of Comparative Labour Relations 53, 53-80 (Roger Blanpain et al., eds. 2017).The Notion of Employer in the Era of the Fissured Workplace - Should Labour Law Responsibilities Exceed the Boundary of the Legal Entity? 53 – 80 (Wolters Kluwer 2017).
  22. [22] See, for example, Commonwealth, A Report of the Fair Work Ombudsman’s Inquiry into 7-Eleven (April 2016).
  23. [23] Fair Work Amendment (Protecting Vulnerable Workers) Act 2017 (Cth), passed by Parliament on September 5, 2017.
  24. [24] Parliament of Queensland, Finance and Administration Committee, Inquiry into the practices of the
labour hire industry in Queensland, Report No. 25(June 2016); Parliament of South Australia, Final Report: Inquiry into the Labour Hire Industry, Report No. 93 (October 2016). Forsyth, supra note 8.
  25. [25] Hardy, supra note 15.
  26. [26] Forsyth, supra note 8.
  27. [27] See amongst others: Employers “pulling hair out” over new labour hire scheme, Workplace Express (May 10, 2019).
  28. [28] See table prepared by Anthony Forsyth, Overview of Labour Hire Licensing Legislation in Queensland, South Australia and Victoria, Corrs Chamber Westgarth (June 26, 2018), available at https://corrs.com.au/site-uploads/images/PDFs/Insights/article-employment-labour-Victorian-Labour-Hire-Licensing-Act-passed.pdf.
  29. [29] Labour Hire Authority, https://labourhireauthority.vic.gov.au/; Labour Hire Licensing Queensland, https://www.labourhire.qld.gov.au/i-provide-labour-hire/laws-and-compliance.
  30. [30] The maximum amount of penalties for engaging in prohibited conduct in Victoria is approximately $ 140,000 for natural persons, and approximately $ 500,000 for corporations. In Queensland it ranges approximately between $ 130,000 and $ 400,000. See Labour Hire Licensing Act 2017 (QLD) s 11, 12, 13; Labour Hire Licensing Act 2018 (VIC) s 13, 14, 15, 16, 94.
  31. [31] Labour Hire Licensing Act 2018 (VIC) s 7; Labour Hire Licensing Act 2017 (SA) s 7. The Explanatory Memorandum of the Victorian legislation clarifies that determining whether the workers are working for the business of the host (labour hire) rather than for the business of the provider (subcontracting) requires going through a series of factors. Control exercised by the host is not determinative. Other factors that need to be taken into account are low skilled/low paid activities, integration in the host’s business, or the fact that the activity used to be performed by the host’s direct employees.
  32. [32] Id
  33. [33] Labour Hire Licensing Act 2018 (VIC) s 8.
  34. [34] Labour Hire Licensing Act 2018 (QLD) s 7. Note that the Queensland Licensing Act does not specify “in and as part of the business . . . of the other person” and therefore could potentially extend to a broader set of arrangements, in which workers are supplied to perform work for an individual that does not run a business. However, information provided by the relevant regulatory authority appear to exclude genuine subcontracting.
  35. [35] Labour Hire Licensing Act 2018 (VIC) s 9.
  36. [36] Labour Hire Licensing Queensland, See current information provided on the website of the Queensland authority under ‘What is not a labour hire service?’ at https://www.labourhire.qld.gov.au/i-provide-labour-hire/licensing; Labour Hire Authority, the website of the Victorian authority, General definition of labour hire services at https://labourhireauthority.vic.gov.au/provider/general-definition-of-labour-hire-services/.
  37. [37] Labour Hire Licensing Regulations 2018 (QLD) s 4.
  38. [38] See Labour hire laws, South Australian Government Consumer and Business Services (June 14, 2019), available at https://www.cbs.sa.gov.au/campaigns/labour-hire-laws.
  39. [39] Labour Hire Licensing Regulations 2018 (VIC) s 4.
  40. [40] Labour Hire Licensing Regulations 2018 (VIC) s 5-6.
  41. [41] According to the Greek mythology, killing the Hydra of Lerna was the second of the twelve labors of Hercules. The myth narrates that when one of the Hydra’s heads was cut off, two more grew in its place.
  42. [42] Maria Azzurra Tranfaglia, Agency Work and its Regulatory Challenges: Lessons Learnt Through a Comparative Overview of Australian and Italian Approaches, in The Evolving Project of Labour Law (Howe, Chapman, Landau eds. 2017).
  43. [43] See Andrew Stewart & Tess Hardy, Submission to Senate Inquiry into the Exploitation of General and Specialist Cleaners Working in Retail Chains for Contracting or Subcontracting Cleaning Companies, available at https://www.aph.gov.au/Parliamentary_Business/Committees/Senate/Education_and_Employment/ExploitationofCleaners/Submissions (2018).
  44. [44] See Woolworths Group, Statement of Principles, available at https://www.woolworthsgroup.com.au/content/Document/Woolworths%20Group%20Statement%20of%20Principles_FINAL.pdf.
  45. [45] See Anna Patty, Second class citizens': ALP pledge to lift pay for labour hire workers, The Sydney Morning Herald (July 16, 2018), available at https://www.smh.com.au/business/workplace/second-class-citizens-alp-pledge-to-lift-pay-for-labour-hire-workers-20180716-p4zrqp.html.
  46. [46] See Anthony Forsyth, Our 'culture of underpayment' must be eradicated, The Sydney Morning Herald (April 5, 2019), available at https://www.smh.com.au/business/workplace/our-culture-of-underpayment-must-be-eradicated-20190319-p515e2.html.


Dispatch No. 17 – Netherlands – "Contradictory Court Rulings on the Status of Deliveroo workers in the Netherlands," by Nuna Zekic*

July 1, 2019

I. Introduction

On January 15th, 2019, two verdicts were issued regarding Deliveroo workers in the Netherlands. One was on the classification of Deliveroo riders as employees,[1] and the second on the application of the collective agreement “Road transport and haulage by road” on meal delivery by Deliveroo riders.[2] Both cases were commenced by Federatie Nederlandse Vakbeweging (FNV), the largest trade union in the Netherlands. The verdict on the employment status is especially important, because half of a year earlier, the same court ruled in a case commenced by one individual Deliveroo rider that he in fact was an independent contractor and not an employee.[3] Why do courts in comparable situations come to such contradictory rulings?

II. The Employment Contract in the Netherlands

The Netherlands does not have a specialized labor court, like some other countries do; instead, regular civil law courts deal with labor and employment cases. The employment contract is defined in the Dutch Civil Code as a contract whereby employees commit themselves to perform labor in service of an employer in exchange for remuneration.[4] The part “in service of the employer” implies that the employee is working under control of the employer. When there is no obligation on the side of the worker to perform labor, Dutch labor law scholars will often decide that there is no employment relationship.[5] As in other systems, the control element is considered to be the characteristic feature of the employment contract. The three elements—labor, remuneration, and control—are important to determine whether a worker is indeed working under an employment contract, but they can be present in other working arrangements too, such as contracts with independent contractors.

Freedom of contract is important in Dutch labor law. The Dutch Supreme Court finds that in principle, the parties in the labor market can contract for work under (many) different agreements. What is applicable between parties is determined by what they had in mind when concluding the agreement, while also taking into account how they actually performed the agreement.[6] That means that the intentions of the contracting parties are very relevant for the classification of the contract, as well as the conduct of the parties. In fact, all of the relevant facts and circumstances of the case have to be taken together, without a single circumstance being decisive. This is called the “holistic approach,” which is the leading approach in Dutch jurisprudence. In this approach, the social status or the social position of the worker can also be relevant: the more economically dependent the worker is, the likelier it is that the judge will rule the worker to be an employee.[7]

The employment qualification test has only gained in importance in recent years. In the Netherlands, workers can only be classified either as employees or as independent contractors (e.g. solo self-employed workers). However, the increase of the use of nonstandard types of employment contracts has been considerable in the Netherlands in the past decade.[8] Flexible employment as well as the number of solo self-employed workers are rising. In 2015, 21% of the workers had a flexible contract, while in 2003, this was 13.6%.[9] It seems that especially the most flexible and insecure types of employment contracts, such as zero-hours contracts, are on the rise.[10] Moreover, transition from flexible work to more permanent positions have decreased.[11]

III. Deliveroo Rulings

Deliveroo has been active in the Netherlands since 2015. Initially, Deliveroo concluded employment contracts with the riders, but from 2017 it stopped using employment contracts and started concluding agreements with riders as independent contractors. The judge in the latest case (the FNV-ruling) therefore chose as the central question whether the character of the legal relationship between Deliveroo and its riders has changed in such a way that the elements of the employment contracts—notably the element of subordination—are no longer present. In the first Deliveroo ruling the judge did not consider this fact. Both judges use the above “formula” established by the Dutch Supreme Court in which not only the rights and duties are considered that the contracting parties had intended to agree, but also the way in which they actually executed their agreement.[12]

In the first Deliveroo ruling, the judge stayed close to the wording of the contract and emphasized that the contract explicitly mentioned that the parties did not intend to enter into an employment contract. Furthermore, the judge found that the rider was aware that he would be working as an independent contractor, since he asked for this contract himself and since he had registered as independent contract at the Chamber of Commerce. In the FNV-ruling, however, the judge chose to emphasize that the contract used by Deliveroo was a standard type of contract that was completely and unilaterally formulated by Deliveroo and that this contract was in fact nonnegotiable.[13] In such situations, the written contract could not be of overriding importance in determining the intention of the contracting parties. The court remarked that it understood the need for flexibility on the side of Deliveroo and even on the side of some riders, but it ruled that the parties cannot decide to opt out of labor law given its imperative and protective nature.[14] The court then considered the meaning of “independent contractor” and decided that a certain element of entrepreneurship must be present—something this judge found to be missing in Deliveroo riders. On the contrary, the work performed by the riders formed the core of the Deliveroo business. Even though Deliveroo stressed the technology aspect of the company and the fact that it was exploring and developing other possible markets, delivery of meals was an essential part of (the identity of) the company; something that was reflected in the name as well.[15] In sum, in the FNV-ruling, the judge paid less attention to the contract and the intentions of the contracting parties. This is where this ruling differs from many other court rulings in the Netherlands. 

As previously stated, the judge examined whether the legal relationship between Deliveroo and its riders changed in such a way that the elements of the employment contract were no longer met. Since Deliveroo riders were first employees and then later were labeled independent contractors, the judge assessed how much their work and the working relations with Deliveroo changed in the meantime. Notably, the judge found that the contractual obligation to accept and perform work changed considerably after the riders became independent contractors. When they were employees, they were obliged to be available for work during a certain minimum of time and they were obliged to accept and perform the offered “rides” during their shift.[16] Declining deliveries (time and again) would result in a summary dismissal. In contrast, there was no obligation to be available for work and to perform work for Deliveroo riders as independent contractors. They were free to decide if and when they logged in the app. The obligation to perform work only arose when the rider accepted a delivery in the app. Notwithstanding this fact, the judge stressed that it was important to assess whether the obligation to be available for work really changed when one looked at the actual situation in practice. The judge assumed that it was not plausible that Deliveroo and the rider signed the contract for incidental delivery. The small remuneration that the rider received for one delivery compared to the costs of purchasing the meal box and the entry in the register of business names at the Chamber of Commerce were signs that the rider did not intend to work for Deliveroo only incidentally. This was also not Deliveroo’s intention, the judge found. The system “Frank” that Deliveroo uses to assign deliveries to riders works in such a way that riders who have applied for shifts in advance are given priority. That means that Deliveroo prefers that riders log in and are available for shifts.[17] Deliveroo also monitors riders’ performance. Riders who deliver often and well can also get “priority access” to certain shifts and areas, which increases their chances of getting assigned deliveries in popular shifts. There is also a bonus system that encourages the riders to work as much as possible and to perform the work well. If the rider wishes to generate enough income, it is in their best interest to log in for a shift and to perform the work well. In addition, if the rider declines a delivery offer, they need to fill in a reason for the decline. Deliveroo can use this—even when the rider fills in no reason—to measure performance. Overall, the judge found that there was no complete freedom to decide whether or not to (be available for) work on the side of the riders. Deliveroo tried to convince the judge otherwise and presented a monthly survey of all deliveries, which shows that 44.5% of all deliveries offered to riders were in fact declined. This was not enough to change her mind, however, since the judge could not conclude from this survey why the deliveries were declined or what were the consequences of declining deliveries.

Deliveroo also referred to the fact that the riders are not obliged to perform the work in person. The contract states that they are free to find a replacement, as long as the substitute has shown beforehand a proof of identity and a permit to work in the Netherlands.[18] However, the judge found this argument to be of little substance. There was very little time to find a replacement when one accepted the delivery, so the need and the ability to make use of this contractual possibility were small. The law indeed states that the employee is obliged to perform the work in person; in performing their work they can only be replaced by a third party with the employer’s permission.[19] The judge found this not to be a distinguishing criterion. According to Dutch law, a subcontractor is obliged to perform work in person as well, unless otherwise agreed.[20] Besides, the possibility of replacement in Deliveroo’s contract can be seen as permission given by the employer.

The judge concluded that when the riders are at work, there is unabated control. It may be the case that there are only general instructions on how to perform the work and that there are no concrete instructions every time the rider needs to make a delivery, but with such “simple, unskilled standard tasks,” no additional instructions are needed. It is sufficient to give general instructions.[21] There is a clear contradiction here with the first Deliveroo ruling from July 2018 where the judge placed these general instructions under the heading of “safety” and remarked that such instructions are not “illogical” in view of the nature of the service. In the context of reliability and safety, Deliveroo was allowed to impose such rules on “those who perform work under its flag.” Such (safety) rules could be given both to independent contractors and to workers. That distinction was not relevant in this case according to the judge, since the rules were rather related to the nature of the service and not so much to the relation between Deliveroo and the riders.[22] In addition, the judge clearly remarked that these were general rules and not case-specific. 

 IV. Analysis

In the FNV ruling, the judge declared that the central question was whether there was (still) a subordinate relationship between Deliveroo and its riders. The main arguments were that real freedom was missing and that the possibility of replacement was an empty shell. The judge stressed the underlying mechanisms, which restricted the real freedom of riders to decide when and if to work for the platform, such as low salaries and the consequences of rejecting shifts.

The Dutch judges clearly disagree on the question of whether or not Deliveroo exercises control over how the riders perform their work. Both see that Deliveroo only gives general instructions, but for one judge (in the FNV case), this is enough to establish control because the type of work does not require other instructions. The other judge that ruled that Deliveroo riders are self-employed found that these general rules are related to and necessary for the type of service Deliveroo is providing and that they do not indicate an employment relationship. The judge in the FNV ruling stressed that GPS control offers a possibility for Deliveroo to follow and monitor the riders.

One of the important aspects in such cases is whether or not the riders are required to use a certain uniform or a meal box and whether or not they have to operate under the logo of the platform. Through the Dutch cases, it is clear that Deliveroo has changed its strategy on this point, at least in the Netherlands. In the first ruling—July 2018—it was not clear whether the use of Deliveroo material was mandatory. In the second ruling—January 2019—it was clear that since Deliveroo stopped using employment contracts, they also no longer require riders to use their material. This change of strategy was not enough to convince the judge in the second case that there was no subordination. Indeed, there still might be strong incentives for the workers to use the uniforms and the meal boxes because of the convenience and the need to be recognizable as meal deliverers both in traffic and with consumers.

In sum, the Dutch test for employment status takes a holistic view: all the circumstances of the case need to be considered in relation to one another. The Supreme Court explicitly mentioned the social status of the parties as one of the circumstances that can be considered. In this context, the economic dependency of the worker in question plays an important role as well. In other words, the judge will often examine whether the worker truly and deliberately chose to be an independent contractor. However, not one factor is decisive in this test. The consequence of this holistic view is that very similar cases can have very different results. Courts can come to different conclusions even in cases where the same employer and the same type of work is concerned. This seems to be the result, at least in part, of differences in the importance given to different elements of the test to determine employment status.

*Nuna Zekic is Associate Professor at Tilburg University, Department of Labour Law and Social Policy.


  1. District Court Amsterdam 15 January 2019, ECLI:NL:RBAMS:2019:198.
  2. District Court Amsterdam 15 January 2019, ECLI:NL:RBAMS:2019:210.
  3. District Court Amsterdam 23 July 23 2018, ECLI:NL:RBAMS:2018:5183.
  4. Art. 7:610 BW.
  5. E.g., Willem H A C M Bouwens, Rogier A A Duk & Wim C L van der Grinten, Arbeidsovereenkomstenrecht 4 (2018).
  6. Dutch Supreme Court 14 November  1997, ECLI:NL:HR:1997:ZC2495 (Groen/Schroevers).
  7. Bouwens, supra note 6, at 11.
  8. Lian Köster & Wendy Smits, Tijdelijk Werk: Nederland in Europees Perspectief, in Dynamiek op de Nederlandse Arbeidsmarkt 129, 147-48 (2014); Arjan Heyma & Siemen van der Werff, De Sociaaleconomische Situatie van Langdurig Flexibele Werknemers, Seo Economish Onderzoek (2013).
  9. Jonneke Bolhaar, Arne Brouwers & Bas Scheer, De flexibele schil van de Nederlandse arbeidsmarkt, CPB Achtergronddocument (Nov. 17, 2016), available at https://www.cpb.nl/sites/default/files/omnidownload/CPB-Achtergronddocument-17nov2016-De-flexibele-schil-van-de-nederlandse-arbeidsmarkt-een-analyse-op-basis-van-microdata.pdf; Other documents report an increase from 16% in 2003 to 27% in 2017, Statistics Netherlands (CBS), Werkzame beroepsbevolking; positie in de werkkring (Mar. 14, 2019), available at https://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=82646ned .
  10. Bolhaar, supra note 10.
  11. Köstersupra note 9;. see also Ronald Dekker, Doorstroom van Flexwerkers, ESB Arbeidsrecht 70, 70-73 (2012).
  12. Dutch Supreme Court supra note 7; Dutch Supreme Court 25 March 2011, ECLI:NL:HR:2011:BP3887 (Gouden Kooi).
  13. District Court Amsterdam 15 January 2019, ECLI:NL:RBAMS:2019:198, [21].
  14. Id. at 22.
  15. Id. at 25.
  16. Id. at 29.
  17. Id. at 31.
  18. Id. at 38.
  19. Art. 7:659 BW.
  20. Art. 7:404 BW.
  21. District Court Amsterdam supra note 2, at 54.
  22. Id. at 20.


Dispatch No. 16 – Spain –"The Next Spanish Labor Law Reform: The Spanish Government has Reached an Agreement with the Trade Unions—but not with the Employer Associations—to Soften Key Aspects of its Labor Law Reforms," by Adrián Todolí Signes*

April 6, 2019

I. Context of the Agreement

In June 2018, there was a change of government in Spain following a vote of no confidence in the president of the conservative Partido Popular. This vote of no confidence, lodged by Pedro Sánchez of the left-wing PSOE party, obtained the majority vote needed for him to become the president of the Spanish government. This circumstance is relevant to the analysis of the agreement for two reasons: i) to highlight the fact that this new government does not have four years to implement its electoral program but only the two remaining years of the term in office, which places it under pressure to demonstrate its capacity to change things with respect to the previous government that it replaced after the vote of no confidence; and ii) to point out that there are no new parliamentary majorities after an election, but instead that the change of government was made possible by the fact that some parliamentary parties that previously supported the conservative government now support the new one. This second point is of great importance since it implies that the new government is aware that, although it has the parliamentary backing it needs to govern, it may not have enough to pass certain measures affecting labor relations.

On the other hand, it must be noted that, in 2012, the previous conservative government passed a comprehensive labor reform package that modified many of the individual and collective rights of Spanish workers. This reform was criticized by the trade unions[1]—and by much of the doctrine[2]—for the following reasons: i) it weakened collective bargaining; ii) it increased the employer's capacity to unilaterally determine working conditions; iii) it increased the employer's ability to unilaterally modify working conditions without sufficient justification; iv) it facilitated collective dismissals without sufficient justification; v) it reduced the amount of compensation due for the unjustified dismissal of workers; vi) it increased the employer's capacity to unilaterally set working hours and overtime for workers; and vii) it increased fixed-term employment with the creation of new types of work contracts.

This reform was implemented unilaterally by the government, without reaching any agreement with trade unions or with employer associations but was applauded by the business confederation for introducing greater flexibility in labor relations.[3]

II. Matters Included in the Current Agreement between the Government and the Major Trade Unions

The agreement discussed here was approved on December 13, 2018 between the Spanish government and the main Spanish trade union confederations—CCOO and UGT. In it, the government undertakes to modify by legislative means the aspects that are considered—according to the government and trade unions—most detrimental for workers in the 2012 labor reform. However, the agreement does not specify a deadline for the amendment and neither does it state exactly what legal instrument will be used to implement the changes. The legislative modifications that have been proposed are the following:

  • 1) Re-introduce ultra-activity into collective bargaining agreements
  • 2) Establish priority in the application of the collective bargaining agreement of the sector over that of the company
  • 3) Oblige subcontractors to pay the same salary or apply the same collective agreement to their workers as the main company
  • 4) Establish the obligation for the company to record the effective working hours of its workers

A. Ultra-activity of Collective Bargaining Agreements

In Spain, this agreement remained in force in the part relating to employees' working conditions (ultra-activity of the collective bargaining agreement art. 86.3 ET) until the 2012 labor reform, when a fixed-term collective bargaining agreement came to an end. That is to say, even if a collective bargaining agreement with a term of three years had expired, it remained in force, as required by law, until there was a new agreement to replace it. The Spanish model of industrial relations entrusts the regulation of working conditions mostly to the collective bargaining agreements, while the Statute of Labor clearly plays a secondary role in regulating labor conditions. This is why, until the 2012 reform, it was understood that there was a need to guarantee the labor conditions of workers between the expiry of one collective agreement and the approval of the next one.

The 2012 reform, however, put an end to this ultra-activity by establishing that once the collective bargaining agreement expired it would only remain active for one additional year. This meant that employers only had to wait one year before being able to stop applying the working conditions guaranteed by the collective agreement. It also put trade unions under strong pressure to accept working conditions that were worse than those in the previous collective bargaining agreement.

Well, the first proposal of the new agreement between the trade unions and the government returns to the previous situation by re-introducing ultra-activity sine die into collective bargaining agreements; that is, the collective agreement remains in force until another agreement is approved to replace it.

B. Priority in the Application of the Sectoral Collective Agreement

In many European countries the issue of centralized and decentralized collective bargaining is the subject of much debate. Spain is no exception. Prior to the 2012 labor reform this debate was resolved by leaving the social partners to reach a decision as to which type of collective agreement had priority: the sectoral agreement or the company agreement.[4]

With the 2012 labor reform, however, priority in the application of the company agreement became mandatory.[5] In other words, trade unions and employer associations cannot, even by agreement, establish a priority other than that of the company agreement. This has led to the creation of many companies with their own agreements that establish wages well below those of the sector. Subcontracting has also increased, as some companies prefer to outsource part of their production to companies that have company agreements. It has also given rise to strong competition among companies based on low wages and not on the quality of their products and services.

This agreement aims to centralize collective bargaining again, leaving the priority in the application of the agreements up to the social partners.

C. Oblige Subcontractors to Pay the Same Salary or Apply the Same Collective Agreement to their Workers as the Main Company

Subcontracting and the decentralization of production have increased greatly in Spain in recent years. Many companies have dismissed workers and then immediately engaged the same services through a third company that pays those workers lower wages. This has meant that at present the decentralization of production is not being carried out due to reasons related to specialization (because the subcontractor company is a specialist in that field) or to the quality of the services provided by the third company, but instead the work is outsourced simply as a way to save labor costs.

The agreement between trade unions and the government proposes that the subcontractor company should pay its workers the same salary as the main company pays to its own workers. This regulation is similar to the one that already exists for Temporary Employment Agencies. Indeed, according to Directive 2008/104/EC of the European Parliament and of the Council of 19 November 2008 on temporary agency work,[6] workers employed through a temporary agency have the same rights as workers in the main company.

The main difference between this regulation and the one proposed in the agreement is that the latter is not as ambitious. Thus, there is no principle of equal treatment between the employee of the main company and that of the subcontractor company, but only equal treatment in terms of pay. In short, the aim of both regulations is precisely to prevent subcontracting from being carried out with the sole purpose of saving labor costs.

D. Establish the Obligation for the Company to Record the Effective Working Hours of its Workers

This last point in the agreement aims to resolve a problem concerning unpaid overtime in Spain and of fraudulent conduct in part-time contracts. Indeed, in recent years, both the trade unions and also the labor inspection authorities have been reporting the existence of employees with part-time contracts who in fact work full-time. In this way, the company pays less tax and lower Social Security contributions, which eventually has a detrimental effect on the worker's retirement benefit.

Labor inspectors also complain that they do not have sufficient tools to address this kind of fraud. The proposal of the government and the trade unions is to oblige all companies to keep an effective record of the hours worked by employees. This would allow the inspector to check whether the employee has worked more hours than he or she should have done. This proposal has been taken up and accepted by the trade unions and incorporated into the agreement.

III. Conclusions

Firstly, it can be seen that much of the 2012 labor reform remains unchanged under this agreement. In fact, the agreement does not propose to increase the amount of compensation paid for dismissal of workers or to reduce the possibilities of the company legally carrying out collective dismissals with little justification. No modification is made concerning the existing possibilities of the employer unilaterally altering working conditions, and no effective formulas are put forward to reduce seasonality in Spain.

The second conclusion that can be drawn is that proposals for labor reform are being made that are not directly aimed at reversing the 2012 reform, although they do combat its detrimental effects on workers. This is the case, for example, of the negative effects that the 2012 reform caused in terms of flexible working hours in favor of the employer and abusive practices in the part-time engagement of workers.

The third conclusion that can be drawn is that the agreement does not specify a deadline for its approval. The government has undertaken the approval of these measures as soon as possible, but without setting a deadline. At present, the government must obtain sufficient parliamentary support to pass these reforms to which it has committed itself with the trade unions.

*Adrián Todolí Signes is Professor of Labor Law and Social Security at the University of Valencia. Licenciado en Economía y Derecho.


  1. Comisión Ejecutiva Confederal, Efectos de la Reforma Laboral  (Sept., 2013), http://www.ccoo.es/bffd07942760f1f51e6438e745df2e3f000001.pdf 2200112.
  2. Jesús Cruz Villalón, Impacto de las Reformas Laborales sobre la Negociación Colectiva, 120 Temas Laborales: Revista andaluza de Trabajo y Bienestar Social, 13-32 (2013) (Spain). Jesús Cruz Villalon, Efectos Incontrolados de la Reforma (El País, 2012), https://elpais.com/economia/2012/02/12/actualidad/1329009978_726310.html.
  3. José María Lacasa, La Reforma Laboral Desde el Punto de Vista Empresarial, 68 Boletin de Estudios Económicos 191-204 (2013) (Spain).
  4. The concept of priority refers to which agreement should be applied to an employee when there are two agreements in the same domain: one at a higher, multi-company level and the other within the company itself.
  5. This priority in the company agreement is limited to certain matters, which are listed in art. 82.3 ET. What happens is that this list includes the most important matters in a collective bargaining agreement, such as wages, which makes it practically an absolute priority. See Eva Lopez-Terrada & Adrián Todoli-Signes, El Convenio Colectivo de Empresa en Materia Salarial: Un Estudio sobre la Prioridad Aplicativa y sus Problemas Jurídicos, 7 Derecho de las Relaciones Laborales 743-753 (2015) (Spain).
  6. 2008 O.J. (L 104) 9.


Dispatch No. 15 – Iceland & Russia – "To Protect the Right to Strike or Not? The Question Before the European Court of Human Rights in app no 2451/16 Association of Academics v Iceland and app no 44873/09 Ognevenko v Russia, by Tonia Novitz*

March 29, 2019 


In 2018 two differently constituted Chambers of the European Court of Human Rights (ECtHR) encountered two quite distinct cases that raised the question as to whether protection of the right to strike was required under Article 11 of the European Convention for the Protection of Human Rights and Fundamental Freedoms (ECHR). 

The contrast between the cases and their outcomes seems stark. The first, Association of Academics v Iceland[1] concerned the extent to which the state can intervene in otherwise legitimate trade union activity in the context of a genuine industrial dispute within the health sector. Was legislative imposition of compulsory arbitration acceptable? To which the ECtHR provided an emphatic answer in the affirmative on the basis that the state’s conduct was not disproportionate and, in any case, comfortably fell within a wide margin of appreciation (as had already been stated by the Icelandic Supreme Court, which had balanced the relevant interests engaged here). There was no mention in the case of International Labour Organization (ILO) standards and the ECtHR drew heavily on the controversial findings of the ECtHR in National Union of Rail, Maritime and Transport Workers v UK.[2] The second case, Ognevenko v Russia[3] was not so much related to the collective rights of a trade union but the right to strike of an individual worker. In finding a breach of Article 11 of the ECHR, the majority judgment made extensive reference to “ILO principles concerning the right to strike,” which state that railway transport is not an essential service unless exceptional circumstances arise, regarding which the Russian state had provided no evidence. Accordingly, Russian legislation that restricted the train driver Ognevenko’s ability to take industrial action, ultimately leading to his dismissal, was not considered “necessary in a democratic society,” but rather disproportionate. Notably, it was observed that the Russian courts had no opportunity to engage in any balancing of relevant interests being required only to ensure formal compliance with the Russian legislation, which they had done emphatically. Pecuniary and nonpecuniary damages were therefore awarded in addition to costs and expenses.

It is suggested here that it is possible to reconcile the cases in a number of different ways. A mere juxtaposition of collective trade union versus individual worker rights seems unsatisfactory, but arguably does reveal a tension between two strands of ECtHR case law, which revert to the Court’s natural bias towards individual protections that have been the subject of critical comment. A more convenient explanation for the Council of Europe, seeking to uphold the consistency of ECtHR decisions, might be the very different scope for judicial balancing of public policy objectives in the two cases. Overall, however, I shall argue here that the approach taken by the Chamber of the Court in Ognevenko is preferable insofar as it makes detailed reference to the significance of ILO standards and supervisory findings. Moreover, it is just possible that Association of Academics could have been decided in the same way without recourse to a wide margin of appreciation, but rather with reference to health and safety concerns identified as relevant by ILO supervisory bodies.    

I. Two Strands of Case Law: Collective and Individual

There are arguably two significant strands of case law regarding the scope of the guarantee of freedom of association under Article 11 of the ECHR that relate to protection of the right to strike. Arguably, the unanimous Association of Academics case reflects the first of these regarding the claims that may legitimately be made by trade unions; while the majority judgment in Ognevenko reflects concern with protection of the individual worker as striker from criminal penalties and even disciplinary action by an employer, including dismissal. Arguably, the case law regarding the first strand of case law is less emphatically protective of a right to strike than the second. 

For some considerable period of time, neither collective bargaining nor the right to strike were considered to be rights necessary to the ability to form and join trade unions for the protection of workers’ interests, in accordance with the wording of Article 11(1).[4] However, the negative freedom of an individual worker not to belong to a trade union had been upheld.[5] That approach was revisited in the ECtHR Grand Chamber judgment in Demir and Baykara v Turkey,[6] which found (with reference to Turkey’s international commitments regarding ILO Conventions) that collective bargaining must be regarded as an essential element of freedom of association under the Convention. Moreover, the scope of limitations on the extent to which restrictions could be placed on collective bargaining by civil servants had to be determined by reference to ILO Conventions Nos 98 and 151, which indicated that a wide-ranging ban was inappropriate. That case was brought both by one of the union members and by the President of the union. This may have been a wise maneuver eliciting sympathy both for the individual and the collective entity simultaneously. Subsequently, the ways in which the cases were presented to the ECtHR meant that the two types of applicant tended to part company.

As regards the trade union applicant, hovering in the background to the Association of Academics case was the judgment in UNISON v UK, in which the ECtHR had found (as early as 2002), that the right to strike could be protected under Article 11(1) but was not to be regarded as an essential aspect of freedom of association. On the facts of that case concerning trade union resistance to privatization, interference with the union’s access to industrial action was, curiously, regarded as justified by reference to Article 11(2). That provision requires that any interference with freedom of association under Article 11(1) must be “prescribed by law and . . . necessary in a democratic society in the interests of national security or public safety, for the prevention of disorder or crime, for the protection of health or morals or for the protection of the rights and freedoms of others.” Special exclusions apply in respect of the armed forces, the police and “administration of the state.” In UNISON a wide margin of appreciation was applied, it being sufficient that the rights of others would be protected by the effective delivery of the health service. The observation was also made that there was “no immediate risk of detriment” to the workers in question, as they could take industrial action in the future when their terms and conditions were in fact threatened or downgraded.[7]

The case of Enerji Yapi-Yol Sen marked the first successful claim for interference with the right to strike under Article 11 on the basis of a circular published by the Prime Minister’s Public Service Staff Directorate which, inter alia, prohibited all public sector employees from taking part planned industrial action. The unequivocal nature of this instruction and the conclusion of ECtHR that it did not answer a “pressing social need” indicated disproportionate interference with the applicant union’s rights. However, in National Union of Rail, Maritime and Transport Workers (RMT) v the United Kingdom[8] there were echoes again of the UNISON approach, when the ECtHR stated that: “The Court does not . . . discern any need in the present case to determine whether the taking of industrial action should now be accorded the status of an essential element of the Article 11 guarantee.”[9] Moreover, the UK ban on secondary action at issue in that case had not prevented workers engaging in more limited collective bargaining and a strike, so that it was considered that there had not been sufficient interference with their rights under Article 11.[10]

Subsequently, there has been a more stringent assertion by the ECtHR in a case brought by a Croatian trade union that the strike is “the most powerful instrument to protect occupational interests of its members.”[11] Nevertheless, in the Association of Academics case the Chamber reverted to the position that: “So far the Court has not found that the taking of industrial action should be accorded the status of an essential element of the Article 11 guarantee,” although it is clear that strike action is protected by Article 11 as it is considered to be a part of trade union activity.[12] Moreover, the limited access to industrial action prior to the imposition of arbitration was treated as sufficient. That this did not “lead to the outcome desired by the union and its members does not mean that the exercise of their Article 11 rights is illusory. . . . Nor does the right to strike imply a right to prevail.”[13] On this basis, a broad margin of appreciation under Article 11(2) was deemed appropriate for the Icelandic state given the other interests at issue.[14]

By way of contrast, the approach of Ognevenko v Russia[15] chimes with a series of earlier Turkish cases concerned with the protection of individual striking workers. In Saime Özcan v Turkey[16]and Urcan v Turkey,[17] the ECtHR had found attempts to impose significant fines on striking secondary school teachers in the public sector to be in violation of their Article 11 rights. In Karacay v Turkey[18], Kaya and Seyhan v Turkey[19] and Çerikçi v Turkey,[20] a breach arose from disciplinary action taken against public servants who had each participated in industrial action. Civil suits brought by the Turkish government against toll booth operators for taking three hours industrial action were also found to constitute a breach of Article 11 in Dilek et al v Turkey.[21] These cases seem consistent with the concern expressed in Demir regarding wide-ranging constraints on collective bargaining in the state sector.[22] There has, however, been the apparently anomalous judgment of Trofimchuk v Ukraine,[23] where dismissal of a boiler plant operator was considered justified in the circumstances, but which has been regarded as at odds with a general trend towards individual protection of the human rights of workers before the Court.[24]

In Ognevenko the blanket ban on strikes by railway workers can be considered akin to the total Turkish prohibition on civil servant collective bargaining and industrial action, creating palpable personal injustice to the individuals affected. On this basis, the majority judgment emphasised previous findings that the right to strike was “one of the most important means” by which trade unions and their members could defend their interests and is accordingly protected under Article 11 ECHR.[25] The dismissal (a disciplinary sanction combined with an earlier transgression) therefore amounted to an interference with Ognevenko’s rights as “guaranteed” by Article 11.[26] This interference was “prescribed by law,” but not necessary in a democratic society under Article 11(2) as was evident from ILO and European Social Charter supervisory findings that railway transport is not an essential service, which could endanger the life or health of (a part of) the population.[27] Even if it were an essential service, it would need to be shown that this particular railway driver’s role was essential, given the stringency of the protection derived from Article 11.[28] Nor was any suitable alternative to an all-out ban considered by the state, when a “minimum service” could have provided an alternative form of restriction, indicating disproportionate interference.[29] Dismissal was regarded as a “severe penalty” that was unwarranted, as in the earlier Turkish cases cited in the concluding part of the majority judgment.[30] The only opposition came from the Russian Judge Dedov, staunchly defending his own national position with reference to the wider margin of appreciation given in the RMT judgment,[31] but he failed to persuade the other members of the Chamber. In this sense, the distinction between collective and individual protections seems to be borne out. However, there are other factors that would also seem to have been influential.

II. Deference to “Balancing” of Interests by National Courts

One relevant factor in both the judgments might be the extent to which national courts engaged with the same issues that are of concern to the ECtHR. A key factor in the Association of Academics case was said to be the capacity of the applicant to challenge the Icelandic legislation imposing arbitration before the Icelandic courts. The District Court of Reykjavik considered its legality with reference to Article 74 of the Icelandic Constitution alongside Iceland’s international obligations, including ILO Conventions. When the District Court found for the Icelandic state, appeal was made to the Supreme Court that upheld that judgment.[32] The care and attention to the findings of the courts in the judgment go beyond mere demonstration of exhaustion of remedies. Instead, it is considered pertinent that the Supreme Court “assessed the necessity of the impugned measures” and did so by evaluating “the situation in the hospitals and health care services affecting patient care due to the strike actions by looking at evidence . . . supported by statistical data.”[33] Moreover, the Supreme Court had taken into account the number of meetings conducted previously and the extent of the impasse between the parties.[34] It was in this context that the ECtHR cited again the RMT case to the effect that “although the process of collective bargaining and strike action did not lead to the outcome desired by the applicant’s member unions and their members, this does not mean that their Article 11 rights were illusory.”[35]

By way of contrast, in Ognevenko, Russian courts—in upholding the applicant’s dismissal—had applied legislation prohibiting participation in industrial action on public railways uncritically.[36] The ECtHR found on the facts that the courts’ finding that the applicant was covered by the ban set out in the legislation was neither “unreasonable or arbitrary” and could accordingly be regarded as being “prescribed by law.”[37] However, when considering the Ognevenko’s challenge to the dismissal, the national courts “had to confine their analysis to formal compliance with the relevant Russian laws and consequently could not balance the applicant’s freedom of association with competing public interests.”[38] This limitation combined with the failure of the Russian state to defend its policies seems to have contributed to the finding that there had been a breach of Article 11.

III. The Relevance of ILO Standards and Supervisory Findings

A last relevant factor for determination of these two cases may well be how ILO standards and supervisory findings influenced the ECtHR. The ILO Committee on Freedom of Association (CFA) and the Committee of Experts on the Application of Conventions and Recommendations (CEACR) have been instrumental in recognizing the right to strike as an entitlement flowing from membership of the ILO under its Constitution, and also arising under ILO Convention No. 87. On these bases, the right to strike is understood to be “one of the essential means through which workers and their organisations may promote and defend their economic and social interests.”[39] The ILO supervisory bodies have also elaborated on the legitimate exceptions to a right to strike, which include the entitlement to exclude protection of purely political strikes, but also strikes in “essential services.” Both the Association of Academics and Ognevenko cases concern an attempt by the state to limit strikes in what have been deemed “essential services.”

Controversy over the status of CEACR supervisory findings arose in the ILO, notably after the ECtHR judgments in Demir and Baykara and Enerji Yapi-Yol Sen and prior to the RMT case. The latter refers explicitly to the walkout by the employers’ group from the ILO Conference Committee on the Application of Standards (CCAS) in protest against CEACR observations and direct requests regarding the right to strike.[40] This conflict has, at least temporarily, been resolved by an agreement between the employers and workers group in the ILO endorsed by governments in 2015,[41] so that we might expect return to business as usual within the ECtHR.

Curiously, the judgment of Association of Academics reviews only established ECtHR case law and pays no attention to ILO standards, making only the following nod in that direction: “The degree of common ground between member States of the Council of Europe in relation to the issue arising in the case may also be relevant, as may any international consensus reflected in the apposite international instruments.”[42] This is curious, as ILO standards would have supported the Court’s finding, since ILO supervisory bodies have found it permissible to restrict strikes in essential services “the interruption of which would endanger the life, personal safety or health of the whole or part of the population.”[43] This would accord with the basis on which Association of Academics case is decided, namely the need to maintain health services placed under threat through industrial action.

By way of contrast, the Chamber majority judgment in Ognevenko makes explicit reference to various international human rights standards, including the International Covenant of Economic, Social and Cultural Rights and the European Social Charter, but more significantly (and at length) to “ILO principles concerning the right to strike” derived from the 5th revision of the CFA Digest of Decisions and CFA case law regarding Russia.[44] The reiteration of the CFA findings by the CEACR is also given explicit attention, namely “that the right to strike may be restricted only in respect of public servants exercising authority in the name of the State and in essential services in the strict sense of the term—that is to say services the interruption of which would endanger the life, personal safety or health of the whole or part of the population ” (the basis on which Association of Academics was ultimately decided).[45] Further, the ECtHR referred to the reminder issued by the CEACR to Russia that “railway transport did not constitute an essential service in the strict sense of the term” and that the CEACR “continues to request Russia to ensure that railway workers can exercise the right to strike.”[46] It was on this basis that it was found that while the right to strike is not absolute, exceptions must be “construed strictly” not impairing “the very essence of the right to organise.”[47] ILO supervisory criticism of Russian legislation,[48] ILO norms regarding use of “minimum services,”[49] and ILO concern with dismissal as a sanction,[50] all directed the majority of the Chamber in Ognevenko to finding a breach of Article 11.


Ognevenko then returns to recognition of the significance of ILO standards for the analysis of the ECtHR of compliance with Article 11, in a way that Association Academics could have done without the outcome of that decisions necessarily changing. We have yet to learn whether the apparent distinction between trade union and individual applicants will affect the preparedness of the Chambers of the ECtHR to adopt such an approach. It will also be interesting to see in future cases what influence internal constitutional review may have when compared with the consensus reached in international labor law through ILO standard setting and supervision. These need not and arguably should not be in tension when it comes to protection of freedom of association under Article 11. The Association of Academics case arguably indicates that they have the potential to be mutually reinforcing. However, that supposition has yet to be fully examined in the still inconsistent ECHR jurisprudence relating to the right to strike.

Tonia Novitz is Professor of Labour Law, University of Bristol, UK: tonia.novitz@bristol.ac.uk.


  1. Association of Academics v. Iceland, App. No. 2451/16, 67 Eur. H.R. Rep. SE4 (2018).
  2. National Union of Rail, Maritime and Transport Workers v. U.K., App. No. 31045/10, 2014-II Eur. Ct. H.R 336. See Alan Bogg & K.D. Ewing, The Implications of the RMT Case, 43 Indus. L.J. 221 (2014).
  3. Ognevenko v. Russia, App. No. 44873/09, Eur. Ct. H.R. (Nov. 20, 2018), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-187732%22]}.
  4. See Swedish Engine Drivers' Union v. Sweden, App. No. 5614/72, 1 Eur. H.R. Rep. 617 (1976); Schmidt v. Sweden, App. No. 5589/72, 1 Eur. H.R. Rep. 632, para. 33–36 (discussed in Tonia Novitz, International and European Protection of the Right to Strike 225–31 (2003)).  
  5. Young, James and Webster v. U.K., App. No. 7601/76; 7806/77. 4 Eur. H.R. Rep. 38 (1982), confirmed in Søresnsen v. Denmark, App. No. 52562/99; 52620/99 (Jan. 11, 2006), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-72015%22]}. See also Virginia Mantouvalou, Is There a Human Right Not to Be a Union Member? Labour Rights under the European Convention on Human Rights, in Human Rights at Work: Perspectives on Law and Regulation 439 (Colin Fenwick & Tonia Novitz, eds. 2010).
  6. Demir v. Turkey, App. No. 34503/97, Eur. Ct. H.R. (Nov. 12, 2008), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-89558%22]}. See also K.D. Ewing & John Hendy, The Dramatic Implications of Demir and Baykara, 39 Indus. L.J. 2 (2010).
  7. UNISON v. U.K., App. No. 53574/99 (Feb. 10, 2002).
  8. See National Union of Rail, Maritime and Transport Workers v. U.K., App. No. 31045/10, 2014-II Eur. Ct. H.R 336.
  9. Id. para. 77, 84.
  10. Id. para. 85.
  11. Hrvatski Liječnički Sindikat v. Croatia, App. No. 36701/09, Eur. Ct. H.R. (Nov. 27, 2014), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-148181%22]}, at para. 59. See also id. (De Albuquerque, J., concurring) for an extensive reference made by Judge de Albuquerque to ILO standards, at para. 3.  
  12. See Association of Academics v. Iceland, App. No. 2451/16, 67 Eur. H.R. Rep. SE4 (2018), at para. 24 (citing RMT Workers, App. No. 31045/10, 2014-II Eur. Ct. H.R 336, at para. 77).
  13. Id. (citing RMT Workers, App. No. 31045/10, 2014-II Eur. Ct. H.R 336, at para. 85); see also id. at para. 32.
  14. Id. para. 25.
  15. Ognevenko v. Russia, App. No. 44873/09, Eur. Ct. H.R. (Nov. 20, 2018), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-187732%22]}.
  16. Saime Özcan v. Turkey, App. No. 22943/04, Eur. Ct. H.R. (Sept. 15, 2009), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-93987%22]} (in French).
  17. Urcan v. Turkey, App. No. 23018/04, Eur. Ct. H.R. (July 17, 2008), definitive judgment Oct. 17, 2008, https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-87634%22]} (in French).
  18. Karaçay v. Turkey, App. No. 6615/03, Eur. Ct. H.R. (Mar. 27 2007), definitive judgment June 27, 2007, https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-79902%22]} (in French).
  19. Kaya v. Turkey, App. No. 30946/04, Eur. Ct. H.R. (Sept. 15, 2009), definitive judgement Dec. 15, 2009, https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-93993%22]} (in French).
  20. Çerikci v. Turkey, App. No. 33322/07, Eur. Ct. H.R. (Oct. 13, 2010), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-99946%22]} (in French).
  21. Dilek v. Turkey, App. Nos. 74611/02, 26876/02,& 27628/02, Eur. Ct. H.R. (July 17, 2007), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-81713%22]} (in French).
  22. Demir v. Turkey, App. No. 34503/97, Eur. Ct. H.R. (Nov. 12, 2008), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-89558%22]}.
  23. Trofimchuk v. Ukraine, App. No. 4241/03, Eur. Ct. H.R. (Jan. 28, 2011), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-101310%22]}.
  24. Filip Dorssemont, The Right to Take Collective Action under Article 11 ECHR, in The European Convention on Human Rights and the Employment Relation 359–65 (Filip Dorssemont et al., eds. 2013).
  25. See Ognevenko v. Russia, App. No. 44873/09, Eur. Ct. H.R. (Nov. 20, 2018), https://hudoc.echr.coe.int/eng#{%22itemid%22:[%22001-187732%22]}, at paras. 56–57.
  26. Id. para. 62.
  27. Id. para. 72.
  28. Id. para. 73.
  29. Id. paras. 77–80.
  30. Id. para. 23.
  31. Id. (Dedov, J., disseting), at paras. 19–20.
  32. See Association of Academics v. Iceland, App. No. 2451/16, 67 Eur. H.R. Rep. SE4 (2018), at paras. 10–18.
  33. Id. para. 30.
  34. Id. paras. 31–34.
  35. Id. para. 32.
  36. See Ognevenko, App. No. 44873/09, at para. 40.
  37. Id. para. 65.
  38. Id. para. 82.
  39. Int’l Labor Org. (ILO), Freedom of Association: Digest of Decisions and Principles of the Freedom of Association Committee of the Governing Body of the ILO paras. 522–23 (5th ed. 2006) [hereinafter ILO, Digest]. See also Comm. of Experts on the Application of Conventions and Recommendations (CEACR), Freedom of Association and Collective Bargaining: General Survey 62 (1983); CEACR, Freedom of Association and Collective Bargaining 66 (1994). Note this is also reiterated in a new edition of ILO, Freedom of Association: Compilation of Decisions of the Committee on Freedom of Association para. 752 (6th ed., 2018) [hereinafter ILO, Compilation].
  40. Int’l Labor Conference, ILO, Conference Committee on the Application of Standards: Extracts from the Record of Proceedings No. 19/Part 1, 101st session (2012), discussed in Claire La Hovary, Showdown at the ILO? A Historical Perspective on the Employers’ Group’s 2012 Challenge to the Right to Strike, 42 Indus. L.J. 338 (2013); and Janice Bellace, The ILO and the Right to Strike, 153 Int’l Lab. Rev. 29 (2014). See also National Union of Rail, Maritime and Transport Workers v. U.K., App. No. 31045/10, 2014-II Eur. Ct. H.R 336, at para. 97.
  41. See Tripartite Meeting on the Freedom of Association and Protection of the Right to Organise Convention, ILO 1, 1–4 (Feb. 23-25, 2015), https://www.ilo.org/wcmsp5/groups/public/---ed_norm/---relconf/documents/meetingdocument/wcms_346764.pdf.
  42. Association of Academics v. Iceland, App. No. 2451/16, 67 Eur. H.R. Rep. SE4 (2018), at para. 25.
  43. ILO, Digest, supra note 39, at para. 578. In this scenario a minimum service may be appropriate which is proportionate. See id. paras. 607, 634. See also ILO