Platform work and fixed-term employment regulation

Abstract

Although platform work has been studied by many labour law researchers, mainly the unclear labour law status of platform workers as well as possible new avenues to ensure their protection have been discussed. However, platform work is similar to already-regulated atypical work arrangements and the possibilities of the application of these regulations needs to be analysed. The aim of this article is to analyse the applicability of the Fixed-Term Work Directive (1999/70/EC (FTWD)) to platform workers. The question of whether platform work can be regarded as fixed-term employment according to the FTWD is analysed, and also whether the measures to prevent the abuse of successive fixed-term contracts ensure that platform workers avoid being placed in a precarious position. In the example of four platforms (Uber, Deliveroo, TaskRabbit and Amazon Mechanical Turk), it is argued that many platform workers can be regarded as fixed-term workers for the purposes of the FTWD. The existence of a bilateral fixed-term employment relationship between the platform and the worker can be detected in the case of platforms providing transportation and food delivery services. A bilateral relationship also forms between the client and the worker in the case of platforms providing universal services. In the case of crowdwork, a tripartite temporary agency work relationship forms between the platform, the worker and the user, and the Temporary Agency Work Directive (2008/104/EC), rather than the FTWD, should be applied. The measures foreseen in the FTWD to prevent the abuse of successive fixed-term contracts effectively protect platform workers. Nevertheless, if the Member State only restricts the total period of successive contracts, their working conditions are not improved.

1. Introduction

In recent years legal problems concerning platform work have been intensively discussed among policymakers and researchers. Because of the emergence of digitally mediated platform work, once again the applicability of existing labour law to new forms of work has been tested. Usually platform work is divided into two subcategories. In the case of crowdwork an online platform matches employers and workers, often with larger tasks being split up and divided among a ‘virtual cloud’ of workers. ¹ Services are arranged, provided and paid entirely online. Work-on-demand via app is a form of work in which the execution of traditional working activities (transport, cleaning etc.) is channelled through apps managed by firms that intervene in setting minimum quality standards of service and in the selection and management of the workforce. ²

In the European Union (EU) collaborative platforms, if developed in a responsible manner, have been regarded as a source of jobs and growth. ³ It is hoped that the collaborative economy will lead to new opportunities and new routes into work and may serve as a point of entry to the labour market. ⁴ At the same time, the collaborative economy raises issues regarding the application of existing legal frameworks and blurs lines between consumer and provider, employee and self-employed. This again can create regulatory grey areas. ⁵ The European Parliament (the Parliament) stresses the importance of safeguarding workers’ rights and calls on the Member States and the Commission to ensure fair working conditions and adequate legal and social protection. ⁶

Platform work has also been in the focus of interest of many labour law researchers. The main concern has been the possible precarity of platform workers because of their unclear labour law status. Due to the special features of platform work including its triangular and often temporary nature, relatively high autonomy of the worker in terms of working place and time, and the lack of a common workplace, the determination of the labour law status of platform workers can be complicated. ⁷ Therefore, most of the research has concentrated on assessing whether a platform worker should be classified as an employee or an independent contractor. ⁸ Different options have been provided to improve the labour conditions of platform workers. For example, the adoption of a Platform Work Directive has been suggested, ⁹ the creation of separate category of independent worker has been proposed, ¹⁰ and tailor-made regulation of work-on-demand via app through special employment contracts has also been envisaged. ¹¹ The Parliament, however, has called on the Commission to examine how far existing EU rules can be applied to the digital labour market and ensure adequate implementation and enforcement. ¹² Moreover, it has asked the Commission to publish guidelines on the application of Union law to the various types of platform business models. ¹³

The issue concerning the applicability of existing labour regulation inevitably arises if platform workers are classified as employees, as suggested in a substantial part of labour law research. Also, in many Member States, platform workers have been classified as employees or workers by the courts, ¹⁴ which means that in practice we are already faced with the question of which labour law norms exactly apply to platform workers.

From the beginning the new Directive on Transparent and Predictable Working Conditions ¹⁵ refers to the possibility that the Directive could be applied to platform workers. According to Recital 8 of the Preamble, the interpretation of the Court of Justice of the European Union (CJEU) of the definition of ‘worker’ needs to be taken into account in the implementation of the Directive. Provided that platform workers fulfil the criteria of ‘worker’, they fall within the scope of the Directive.

Platform work also has many similarities with other atypical employment relationships, and it needs to be considered, whether the Directives regulating atypical employment relationships such as the Fixed-Term Work Directive (FTWD), ¹⁶ the Part-Time Work Directive (PTWD) ¹⁷ or the Temporary Agency Work Directive (TAWD) ¹⁸ can be applied to platform workers. So far, research concentrating on this issue is largely missing. Stefano and Aloisi have briefly analysed the applicability of the existing EU labour law legislation to platform work; ¹⁹ Ratti ²⁰ and the author of the present article have discussed the possibilities of applying the TAWD to platform workers. ²¹

This article discusses the applicability of the FTWD to platform workers. More specifically I aim to analyse whether platform work can be regarded as fixed-term employment according to the FTWD, and whether the measures to prevent the abuse of successive fixed-term contracts foreseen in the Directive ensure that platform workers are not placed in precarious positions. Even though the principle of equal treatment is one of the most important provisions of the Directive, I exclude this from my scrutiny - first, in order to study other aspects more profoundly; and second, because in my view it would be more reasonable to study this principle separately in the light of all abovementioned Directives. In this article I also exclude the Directive of Transparent and Predictable Working Conditions from my analysis, although when a platform worker is classified as a ‘worker’ it should be applied in parallel with the FTWD. This is because even though both Directives aim to improve the working conditions of (atypical) workers, different means are used for this purpose. If the FTWD aims to improve the quality of fixed-term work by ensuring the application of the principle of non-discrimination and by preventing the abuse of successive fixed-term employment contracts or relationships, the Directive of Transparent and Predictable Working Conditions tries to achieve this by promoting the transparency and predictability of working conditions. For this reason, I see the measures foreseen in the Directive of Transparent and Predictable Working Conditions as a separate research topic.

I argue that many platform workers can be regarded as fixed-term workers for the purposes of the FTWD. The existence of a bilateral fixed-term employment relationship between the platform and the worker can be detected in the case of platforms providing transportation and food delivery services. The formation of a bilateral fixed-term employment relationship between the client and the worker is possible in the case of platforms providing universal services such as TaskRabbit. In the case of crowdwork, a tripartite temporary agency work relationship forms between the platform, the worker and the user, and the TAWD rather than the FTWD should be applied. The measures foreseen in the FTWD to prevent the abuse of successive fixed-term contracts effectively protect platform workers. Nevertheless, if the Member States only restrict the total period of successive contracts, the working conditions of platform workers are not improved.

I begin with analysing whether platform work can be regarded as fixed-term employment and as such belong to the scope of the FTWD. More specifically, I discuss the existence of an employment relationship in the case of platform work and analyse whether a bilateral or tripartite relationship forms in the case of platform work. I go on to discuss whether the measures to prevent the abuse of successive fixed-term contracts can be applied in the context of platform work and whether those measures can prevent platform workers being placed in a precarious position. I will analyse the formation of successive fixed-term contracts in platform work and the influence of the requirement of objective justification as well as other measures used by the Member States to prevent the abuse of fixed-term contracts to the working conditions of platform workers.

2. Platform work as a form of fixed-term employment

In order to apply the FTWD to platform workers we need to determine whether platform work can be regarded as fixed-term work according to the Directive.

Clause 2(1) FTWD states that ‘this agreement applies to fixed-term workers who have an employment contract or employment relationship as defined in law, collective agreements or practice in each Member State.’ Clause 3(1) sets out that ‘the term ‘fixed-term worker’ means a person having an employment contract or relationship entered into directly between an employer and a worker where the end of the employment contract or relationship is determined by objective conditions such as reaching a specific date, completing a specific task, or the occurrence of a specific event.’

These clauses refer to three important conditions that the platform worker needs to fulfil in order to belong within the scope of FTWD: being in an employment contract or relationship according to the law of the Member State; having a bilateral employment contract or relationship; and having an employment contract or relationship, the end of which is objectively determined.

2.1 Employment relationship

2.1.1 Definition of ‘employment relationship’ in the FTWD

The FTWD explicitly states that national concept of employment relationship is applied to determine the personal scope of the Directive. In most of the countries the main characteristics used to distinguish between employees and self-employed workers are subordination or control, and dependency. ²² Although subordination is still the leading criterion defining the employment relationship, some courts have also started to use the dependency test, that concentrates mainly on the determination of the worker’s dependency (psychological, social, economic) on the employer. ²³ On the basis of the Employment Relationship Recommendation of the International Labour Organisation (ILO) ²⁴ and different legal systems, Davidov argues that 12 relevant indicia can be used to determine the existence of an employment relationship. The first group of characteristics suggests the existence of an employment relationship and the lack of thereof suggests otherwise. Democratic deficit (subordination), including integration into the employer’s organisation, the inability to choose working time and place; an obligation to be available to work; and dependency, including on a single/main employer; the provision of tools and materials by the employer; no chance of profit/risk of loss; no entrepreneurial control; and job specific investments belong to this group. The second group of indicia including direct day-to-day control, the right to weekly rest and annual leave, and a non-competition clause suggests the existence of an employment relationship, but lack thereof does not suggest otherwise. Finally, the absence of continuity of the relationship signifies a lack of dependency and speaks against the existence of an employment relationship. ²⁵

Reference to the use of the national concept of employment relationship in the FTWD means that no EU concept of ‘worker’ should be used. However, in some cases the CJEU has substituted national classification with its own concept of ‘worker’ for the purposes of the application of the rights and principles contained in the Directive. ²⁶

For example, in Adelener the Court ruled that ‘the scope of the Framework Agreement is conceived in broad terms, covering generally “fixed-term workers who have an employment contract or employment relationship as defined in law, collective agreements or practice in each Member State”.’ The CJEU added that ‘the definition of “fixed-term workers”…, encompasses all workers without drawing a distinction according to whether their employer is in the public, or private, sector.’ ²⁷ In del Cerro the CJEU found that the FTWD is ‘applicable to all workers providing remunerated services in the context of a fixed-term employment relationship linking them to their employer’, and highlighted that the Member States cannot remove certain categories of workers from the protection of the Directive by jeopardising its effectiveness and uniform application. ²⁸

In O’Brien ²⁹ the Court did not deal with the personal scope of the FTWD but discussed the personal scope of the PTWD. Nevertheless, the definition of the scope of this Directive is identical to the one of the FTWD, and therefore the reasoning of the CJEU could be used also in the case of fixed-term work. ³⁰ In this case the Court reaffirmed that ‘the concept of “workers who have an employment contract or an employment relationship” was to be interpreted in accordance with national law.’ Yet, the CJEU added that the Member States ‘may not apply rules which are liable to jeopardise the achievement of the objectives pursued by a directive and, therefore, deprive it of its effectiveness.’ ³¹ The Court found that the exclusion from the personal scope of the PTWD may be permitted ‘only if the nature of the employment relationship concerned is substantially different from the relationship between employers and their employees which fall within the category of ‘workers’ under national law.’ Furthermore, the CJEU highlighted an important condition to determine, whether the relationship is substantially different by saying that ‘that distinction must be made in particular in the light of the differentiation between that category and self-employed persons.’ ³²

It can be concluded that as a rule the personal scope of the FTWD is determined according to national law. Nevertheless, the Member States cannot apply rules that jeopardise the achievement of the Directive’s objectives or exempt certain categories of workers from the scope of the Directive unless their relationship is substantially different to that of employers and employees. In the context of platform work, three important observations can be made from this conclusion: first, the Member States cannot exclude the whole category of platform workers from the scope of the FTWD; second, platform workers can be excluded if they classify as self-employed; third, subordinate platform workers are included in the protection of the provisions of the FTWD even if they are not regarded as fixed-term workers or employees for the purposes of national law. It is therefore clear that in order for the FTWD to apply to platform workers, they need to qualify as employees for the purposes of the Directive.

2.1.2 Employment relationships in the case of platform work

The determination of the existence of an employment relationship in the case of platform work is not an easy task. For a start, platform workers have some autonomy as regards their working place and working time, and the conditions of the relationship also differ depending on the platform. Platform workers can be more or less subordinated to the employer. Additionally, three parties participate in the relationship: the worker, the platform and the client, which makes it difficult to determine who is the employer. As Prassl and Risak put it, ‘even if the putative employee has been classified as such, she may struggle to point at her employer, which could be the platform or indeed the crowdsourcer.’ ³³

For the time being the courts in different Member States have classified platform workers as ‘employees’ or ‘workers’ by analysing their subordination to the platform. For example, on 28 October 2016 the Employment Tribunal in the United Kingdom (UK) ruled that Uber drivers are workers, and not self-employed, within the meaning of the Employment Rights Act. ³⁴ On 1 June 2018 the Social Court of Valencia found that the rider of takeaway firm Deliveroo should have been treated as an employee, and not as a self-employed contractor. ³⁵ On 28 November 2018 the Supreme Court in France held that delivery riders working for online delivery platforms such as Take Eat Easy are employees not self-employed workers. ³⁶ On 15 January 2019 the District Court of Amsterdam classified Deliveroo riders as employees. ³⁷

These cases concern transport services and food delivery firms. In order to elaborate the performance of control in these sectors, I describe briefly how the working process is organised. Ivanova et al have studied the performance of control in food delivery platforms. They explain that Deliveroo riders need to swipe ‘Accept’ to accept the order, ‘Arrived’ when they are in the restaurant, ‘Collected’ when food is received, ‘Arrived’ when they are at the customer’s address, and ‘Delivered’ when food is handed over. Riders’ movements and time between the steps is constantly tracked by the GPS. In case of any irregularity, the rider receives a notification through the app. ³⁸

Here, the performance of control by the platform (through the app) is clearly detectable. For example, in classifying Deliveroo riders as the employees of the platform, the Social Court of Valencia found that the control of the work was performed by Deliveroo through GPS. Additionally, the fact that the main means of the production (the app and website) and the brand belonged to Deliveroo; the setting of the price of the service by Deliveroo; the lack of information on the part of the rider as to which restaurants were registered with the Deliveroo app; and the number and volume of orders, inferred the existence of an employment relationship. ³⁹ These facts not only infer that the rider is not a self-employed contractor, but also show that in this relationship the client does not control the working process. The client does not provide tools to perform the work, he/she does not agree the price of service with the rider, and most importantly has no possibility of supervising or directing the work. She/he only orders and receives the food.

Another example concerns transportation. The control performed by Uber is described in the case Aslam et al v. Uber BV et al. ⁴⁰ Passengers register and book their trip by downloading the app and logging in. They may state their destination and on request, can receive a fare estimation. After receiving the request, Uber estimates through their equipment (driver’s smartphone and app) the closest driver to the passenger and informs her/him through smartphone. The driver is told the passenger’s first name and rating and has 10 seconds to accept the trip. After the trip is accepted, driver and passenger are put into direct telephone contact through app. The driver is not aware of the destination until pickup and learns it from the passenger or from the app by pressing ‘Start Trip’ button. The app provides detailed directions to the destination. The driver is not bound to follow the proposed direction and does not if the passenger stipulates a different route. However, if a passenger claims for a refund on the basis that the most efficient route was not chosen, the driver needs to justify their departure from the route indicated by the app. ⁴¹

Although, whilst the client seemingly has a possibility to partly direct the working process by determining the route, the platform can punish the worker for following these directions. Additionally, the platform controls and arranges most of the working process. It recruits drivers, determines the fee, requires the drivers’ vehicles to meet certain specifications, requires drivers to accept and/or not to cancel a certain number of trips and enforces this requirement by deactivating drivers who breach it, and subjects drivers to its rating system that is effectively a performance management and procedure. ⁴² Here, too, the role of the client in controlling the working process is almost non-existent. The client orders the service arranged according to the requirements of Uber and pays the service fee determined by Uber.

Whilst in the abovementioned cases the subordination of the platform worker to the platform and the existence of an employment relationship between those parties can be detected, we cannot automatically broaden this outcome to all platform work. The existence of subordination is determined according to factual circumstances and it cannot be excluded that in some cases the direction and control is performed by the client and the platform only acts as an intermediary by enabling the employers to meet their employees.

For example, TaskRabbit does not perform control over the working process as strongly as abovementioned platforms. TaskRabbit connects people and businesses to get everyday and skilled tasks done. Popular tasks include handyman work, assembling furniture, help with moving, yard work, cleaning etc. ⁴³ A Tasker can be hired the same day or for a future date. In the former, the user does not have the option to choose a particular Tasker. After the user selects a category and provides a description of the job, TaskRabbit sends a notification to all Taskers in the area. The Tasker can choose to accept the task. Once the task is accepted, the user can communicate with the Tasker via phone or the chat function. When hiring a Tasker for a future date, TaskRabbit gives the user the option to choose any Tasker or to select a particular Tasker. If the latter option is chosen, the user is shown a list of Taskers, along with information about each. Taskers can choose to accept or refuse jobs without penalty. In its Terms of Service, TaskRabbit explicitly states that it ‘does not, in any way, supervise, direct, control or monitor a Tasker’s work or Tasks performed in any manner.’ Although users provide ratings for Taskers, TaskRabbit does not investigate this feedback. The Terms of Service include principles that Taskers must adhere to when conducting themselves, but these principles do not directly instruct Taskers on how to perform their tasks. ⁴⁴ Nevertheless, TaskRabbit starts the relationship by activating worker’s accounts and can terminate the account if the Tasker violates the Terms of Service. The platform also organises the payment and participates in setting wage levels. TaskRabbit deletes inappropriate tasks and directs the worker to perform the task personally. ⁴⁵

Prassl and Risak argue that in the case of TaskRabbit the supervision and direction of the work is shared by the platform, the worker and the user. ⁴⁶ Yet they highlight some of the facts that may support the argument that a Tasker can be the ‘employee’ of the user. They explain that work organised through TaskRabbit is highly heterogeneous: some clients provide tools and closely supervise the work, whereas others expect Taskers to bring their own tools. ⁴⁷ TaskRabbit denies any employer role, whilst leaving open the question as to employer status between clients and Taskers. According to the Terms of Service, ‘[e]ach User assumes all liability for proper classification of such User’s workers as independent contractors or employees based on applicable legal guidelines.’ ⁴⁸ Finally, under certain circumstances, Taskers may apply for the longer-term business tasks under the ‘Jobs’ filter. In its blog, TaskRabbit explains that when the Tasker works in a longer-term office job she/he should ‘do as the employees do’ regarding the office etiquette and confidentiality. It is also explained that the Tasker in these jobs is the representative of the business for which she/he works. ⁴⁹

Hence, in the case of TaskRabbit it is possible that the worker can be regarded as the ‘employee’ of the user. This classification, of course, depends on the factual circumstances of the relationship. However, if the user substantially directs and supervises the work, provides tools and materials, determines working time and place, and integrates the worker into the organisation by subjecting her/him to the organisational rules of the user, an employment relationship can be formed between the worker and the user. This is even more likely if the relationship between the user and the worker is continuous, and the worker is bound by confidentiality clauses as in the case of tasks under the ‘Jobs’ filter.

Another platform in which the control of the worker by the user is more substantial is Amazon Mechanical Turk (AMT). Through AMT, workers can perform activities such as copying or translating texts, identifying spelling errors, grouping and labelling items, etc. Each user must register as a Requester or Provider. Requesters post tasks and set compensation and hiring conditions, and can refuse to accept the task while retaining the work done. In this case worker does not get paid. Requesters can also determine the criteria that the worker needs to meet and can define instructions. The platform pays for the work and can suspend or terminate the Provider’s account. ⁵⁰ AMT participates in the selection and management of the workforce and arranges the provision of services and payment through online. It retains the right to terminate the task and de-activate the profile of the Provider. ⁵¹

Even though AMT denies the existence of an employment relationship both between the Provider and Requester, and between the Provider and AMT, ⁵² based on factual circumstances, the worker may be classified as an ‘employee’. Felstiner has analysed the status of the AMT worker in the context of the law of the United States of America. ⁵³ He has detected several factors that speak in favour of the ‘employee’-status of the AMT worker.

Felstiner argues that the AMT worker can be more or less integrated into the organisation of the user. The integration test includes a consideration of the importance of the worker’s work to the business of the beneficiary. The business of some Requesters (such as SpunWrite) is wholly dependent on the work of AMT Providers, while other companies use AMT only periodically or at critical moments. ⁵⁴ The continuity of the relationship also varies depending on whether the same Provider works repeatedly with the Requester or not. However, the open-call system used in AMT discourages continuity. ⁵⁵ The provision of tools and equipment is shared between the Provider and AMT. The provider uses her/his own computer; AMT provides the necessary software and web platform to bring together Providers and Requesters and to enable the completion of tasks, communication, and payment transactions. ⁵⁶ The Provider has control over the place of work and whether to accept tasks; the Requester controls working time through setting the deadlines and directs and supervises the work by giving instructions and refusing the work not completed according to the instructions. ⁵⁷ Finally, the Provider’s opportunity for profit or loss and entrepreneurial control is limited. ⁵⁸

In the case of AMT, too, it is difficult to classify all Providers as ‘employees’ or ‘independent contractors’. This classification, as well as the answer to the question of who should be regarded as the ‘employer’, varies and depends on the circumstances of the concrete relationship. Nevertheless, it seems that the Provider can be classified as the ‘employee’ of the Requester if the business of the Requester is largely dependent on the work of the AMT Providers, the same Provider is used repeatedly, and the Requester gives detailed instructions to the performance of the work.

The aim of this section was not to give a final answer to the question whether platform workers should be regarded as ‘employees’. Because of the variety of platforms and working arrangements organised through platforms, finding a single answer would be difficult. Rather, I aimed to illustrate the different conditions used in platform work. Even though we need to consider all aspects of the relationship in order to determine the employment status of the worker, those examples prove that the existence of an employment relationship between the platform and the worker, or the user and the worker, cannot be excluded. However, this is not sufficient to regard the relationship as fixed-term work. The application of the FTWD also requires the existence of direct relationship between the worker and the employer.

2.2 Direct relationship

According to clause 3(1) of the FTWD, the term ‘“fixed-term worker” means a person having an employment contract or relationship entered into directly between an employer and a worker.’ The Preamble of the agreement sets out that it is applied to ‘fixed-term workers with the exception of those placed by a temporary work agency at the disposition of a user enterprise.’ Therefore, the FTWD cannot be applied to temporary agency workers. This is confirmed also in the practice of the CJEU. In Della Rocca ⁵⁹ the Court held that ‘it is explicitly stated in the fourth paragraph of the preamble to the Framework Agreement that it does not apply to fixed-term workers placed by a temporary work agency at the disposition of a user enterprise, it being the intention of the parties to conclude a similar agreement relating to temporary agency work.’ ⁶⁰

As explained earlier, three parties are involved in a platform work relationship: a worker, a platform and a user. Hence, it is possible that instead of a bilateral relationship, a temporary agency work relationship forms between the parties. These relationships, however, are not covered with the FTWD. Another option is, that this tripartite relationship does not meet to the criteria of temporary agency work, but the platform acts as a placement facilitator, which again does not exclude the application of the FTWD.

The ILO regulates tripartite relationships in the Private Employment Agencies Convention, 1997. ⁶¹ According to Art.1 of the Convention, private employment agency means any natural or legal person, which provides one or more of the following labour market services:

services for matching offers of and applications for employment, without the private employment agency becoming a party to the employment relationships which may arise therefrom;

services consisting of employing workers with a view to making them available to a third party, who may be a natural or legal person (referred to below as a “user enterprise”) which assigns their tasks and supervises the execution of these tasks;

other services relating to jobseeking, determined by the competent authority after consulting the most representative employers and workers organizations, such as the provision of information, that do not set out to match specific offers of and applications for employment.

Services described in Art.1b of the Convention meet the definition of temporary agency work (TAW) used in the TAWD. According to Art.1 TAWD: ‘This Directive applies to workers with a contract of employment or employment relationship with a temporary-work agency who are assigned to user undertakings to work temporarily under their supervision and direction.’ This definition includes two important conditions, first, in the case of TAW, an employment relationship is formed between the agency and the worker; second, the worker works temporarily under the supervision and direction of the user undertaking.

Art.1a of the Convention describes arrangements in which the agency does not become the party of the employment relationship but acts only as the facilitator of the employment relationship between the worker and the user. Countouris and Horton argue that even though these kind of arrangements do not explicitly fall within the scope of the TAWD, some reasons favour their inclusion. They note that often it is difficult to draw a clear line between temporary work agencies and placement facilitators. Also, they refer to Art. 2 TAWD, which sets out that ‘the purpose of the directive is to ensure the protection of temporary agency workers and to improve the quality of temporary agency work…., and by recognising temporary-work agencies as employers…’. Countouris and Horton find that this introduces a strong presumption that in complex cases, the existence of a personal work relationship between the worker and the intermediary should be assumed. They also argue that ‘only where the intermediary retains no link whatsoever with the intermediated worker, should the law accept that it is acting as an “agency” (in the technical sense) and therefore enquire whether the relationship with the user is one of employment or self-employment’. ⁶² Countouris and Horton have analysed the scope of the TAWD in the British context and they argue that the broader category of ‘workers’, not only that of ‘employees’, should fall within the scope of the TAWD. It is also important to keep in mind that in the UK the FTWD applies only to ‘employees’.

It is true that it can be difficult to differentiate between temporary work agencies and placement facilitators, not only because there can be mixed arrangements, but also because the supervision and direction of the work is the main characteristic of an employment relationship. Both in the case of temporary agency work, as well as in the case of placement facilitation, the user directs and supervises the work. It is therefore difficult to determine in which cases the direction and supervision of the work is such that a direct employment relationship is formed between the user and the worker, and in which cases the temporary work agency needs to be regarded as the employer. Broadening the scope of the TAWD to include placement facilitators is not necessarily the best option for the worker. In most of the Member States the TAWD applies to ‘employees’; the application of the TAWD again excludes the application of the FTWD. Both the TAWD as well as the FTWD foresee the principle of equal treatment with directly hired or permanent workers to improve their working conditions. ⁶³ Nevertheless, while the temporary agency worker should be treated equally as regards basic working conditions including pay and working time, ⁶⁴ the fixed-term worker needs to be treated equally in respect of all working conditions. ⁶⁵ Therefore, being regarded as a fixed-term employee of the user can be more advantageous for the worker than of being regarded as the employee of the temporary work agency.

In the context of platform work, the formation of a bilateral relationship is more likely in transportation and food delivery sector. In the case of Uber, as well as Deliveroo, strong direction and supervision of the worker on the part of the platform can be detected. The user has no possibility of directing or supervising the worker; alternatively she/he has only marginal or putative supervisory power. Therefore, no employment relationship forms between the worker and the user. Also, because the direction and supervision of the worker is not assigned to the user, a temporary agency work arrangement cannot be formed. As a result, the platform acts as a sole employer, and the user can be regarded only as the client of the platform. ⁶⁶ As no temporary agency work relationship forms, the worker is directly engaged in an employment relationship with the platform and the FTWD can be applied to those arrangements.

The situation is more complex if the user (partly) directs and supervises the worker. For example, in the case of AMT and TaskRabbit the direction and supervision of the worker by the user can be detected. We need to determine, then, whether the platform acts a temporary work agency or as a placement facilitator. Ratti finds that the significant difference between these two models is that in the case of placement facilitation the role of the intermediary ends when the worker and the user start their relationship; in the case of TAW the intermediary plays an important role during the whole relationship. ⁶⁷ As described above, the role of AMT does not end after the Provider and the Requester have started their relationship. On the contrary, AMT takes care of payment, provides software and the web platform to enable other parties to complete tasks and communicate, and finally retains the right to terminate the task and de-activate the profile of the Provider. Hence, it appears that in the case of AMT, a TAW relationship forms between the platform, the Provider and the Requester, and the TAWD could be applied. The same is concluded by Ratti. ⁶⁸

In the case of TaskRabbit, the platform does not perform as significant a role during the relationship as in the case of AMT. While it is true that TaskRabbit also organises payment and can terminate the relationship, the task is not performed through the web platform of TaskRabbit. The work is performed physically at a time and place determined by the user, using the worker’s or user’s tools, the worker is more or less directed and supervised by the user, and can be integrated into the organisation of the user. Although it cannot be claimed that the platform does not retain any link with the worker, its role does not end at the beginning of the relationship between the worker and the user, and the application of the TAWD is not that straightforward. As the role of the user is more significant than that of the platform, the platform could be considered as the placement facilitator and a direct employment relationship between the worker and the user could be formed.

In conclusion, a direct employment relationship is more likely to form between the platform and the worker in the transportation and food delivery sector. In the case of platforms mediating a range of different physically performed tasks, a direct employment relationship can be formed between the platform and the user. In both cases, the application of the FTWD is possible. In crowdwork arrangements, however, instead of a direct relationship, a tripartite employment relationship, forms. As a result, the TAWD, not the FTWD, should be applied to crowdwork,

3. Preventing the misuse of successive contracts in the context of platform work

Clause 1 FTWD sets out two main aims of the Directive: first, to improve the quality of fixed-term work through the application of the principle of non-discrimination; and second, to prevent abuse arising from the use of successive fixed-term employment contracts or relationships. Clause 5 foresees concrete measures the Member States need to introduce to prevent such abuse. According to Clause 5(1), the Member States need to introduce one or more of the following measures:

objective reasons justifying the renewal of such contracts or relationships;

the maximum total duration of successive fixed-term employment contracts or relationships;

the number of renewals of such contracts or relationships.

Clause 5(2) sets out that the Member States determine under what conditions fixed-term employment contracts or relationships shall be regarded as ‘successive’ and shall be deemed to be contracts or relationships of indefinite duration.

Hence, as a rule the Member States decide which contracts need to be classified as successive, as well as choose whether to adopt one or more of the measures to prevent the abuse of successive contracts. However, the CJEU has not been silent in clarifying the provisions of the FTWD. The Court has evaluated national definitions of ‘successive contracts’ and ruled on the content of ‘objective reasons’. There are no CJEU judgments concerning other two measures (maximum total duration of successive contracts and the number of renewals), which appear to be sufficiently clear so as not to cause interpretation problems. In the following I will discuss whether successive fixed-term contracts can be used in the context of platform work and analyse how the measures provided to prevent the misuse of successive contracts can be applied in this case.

3.1 The use of ‘successive contracts’ in the case of platform work

In Mangold ⁶⁹ the CJEU clarified that Clause 5(1) is supposed to ‘prevent abuse arising from the use of successive fixed-term employment contracts or relationships’ and the interpretation of this provision is irrelevant in the case of the conclusion of a single fixed-term contract. ⁷⁰ On several occasions the Court has reaffirmed that no objective justification is needed if the fixed-term contract is concluded for the first time. For example, in Angelidaki ⁷¹ the CJEU held, that ‘the objective reasons referred to in clause 5(1)(a) thus relate only to the renewal of such contracts or relationships.’ Therefore, the FTWD does not impose on the Member States the obligation to prevent the abuse of single fixed-term contracts; preventive measures need to be taken only if more than one fixed-term contracts are concluded between the same parties.

However, it has been unclear in what case several fixed-term contracts concluded between the same parties can be regarded as ‘successive’. In Adeneler the Court held that the discretion given to the Member States in defining the concept of ‘successive’ is not unlimited and it cannot compromise the objective or the practical effect of the FTWD. The CJEU argued that ‘a national provision under which only fixed-term contracts that are separated by a period of time shorter than or equal to 20 working days are regarded as successive must be considered to be such as to compromise the object, the aim and the practical effect of the Framework Agreement.’ The Court explained that such a restrictive definition would result in the insecure employment of a worker for years, since, in practice, the worker would have no choice but to accept breaks in the order of twenty working days over the course of a series of contracts. The twenty-day rule would also permit the abuse of such relationships by employers and prevent the conversion of fixed-term contracts into stable employment. The CJEU concluded that Clause 5 needs to be interpreted as ‘precluding a national rule,…, under which only fixed-term employment contracts or relationships that are not separated from one another by a period of time longer than 20 working days are to be regarded as “successive”…’ ⁷²

In Fiamingo ⁷³ the Court considered whether fixed-term contracts separated by a time lapse of sixty days can be regarded as ‘successive’. Differently from Adeneler, the CJEU found that ‘such a lapse of time may generally be considered to be sufficient to interrupt any existing employment relationship and to have the effect that any contract signed after that time is not considered to be successive’. The Court argued that it would seem difficult for an employer with permanent requirements to circumvent the protection against abuse afforded by the FTWD by allowing a period of two months to elapse after the end of every fixed-term employment contract. ⁷⁴

The CJEU has clearly prohibited the overly narrow interpretation of the concept of ‘successive contracts’. While the time lapse of twenty days between two fixed-term contracts has been regarded as insufficient to exclude them being ‘successive’, a separation of sixty days means that both fixed-term contracts are regarded as first-time contracts, the abuse of which need not be prevented.

According to Clause 3(1) FTWD, the end of fixed-term contract is determined by objective conditions such as reaching a specific date, completing a specific task, or the occurrence of a specific event. In the case of platform work, work is performed on a task basis. Every task, ride or trip the platform worker takes could be considered as a fixed-term employment contract that begins when the worker starts to perform the task and ends once the task is completed. As platform workers should have autonomy as regards whether, where and when to work, in theory it should be possible that a task is performed every two months. In practice the duration of the task clearly influences the time-lapse between two tasks. The first group of tasks includes very short-term work such as taking a trip with Uber or delivering food with Deliveroo. Also, menial tasks performed through TaskRabbit can belong to this category. If the task is very short-term, it means that platform workers need to perform several tasks in a short period of time in order to earn a reasonable wage. At the same time platforms use different methods, such as sending alerts on opportunities to earn more money ⁷⁵ or warnings of being excluded from the platform if a sufficient number of tasks is not accepted, ⁷⁶ to guarantee that workers perform more work.

The second group of tasks includes longer-term jobs performed through TaskRabbit. In this case it is not impossible that the worker leaves a period of two months between two tasks. Additionally, as already discussed, the issue of who is the employer is not clear in the case of TaskRabbit. If the user is considered as the employer, and TaskRabbit acts only as a placement facilitator, two situations can emerge. On one hand the worker can conclude fixed-term contracts with different users, in which case these are not regarded as successive; on the other hand, more than one fixed-term contract can be concluded with the same user. In the latter case the contracts are not successive if they are separated by a two-month period.

It appears that in the context of platform work, very short-term tasks performed in the transport or food delivery sector often form successive contracts. In TaskRabbit it is possible that despite performing more than one task through the platform, the fixed-term contracts are not successive. Because users and platform can be classified as employers, continuous fixed-term contracts may be concluded with different employers. Additionally, longer-term tasks can be separated from each other by two months. Nevertheless, if work is performed for the same user several times and each instance is separated by less than two months, successive contracts can be formed.

3.2 Preventing the abuse of successive contracts

3.2.1 Objective reasons justifying the renewal

In the CJEU, the content of the first measure, namely the objective reasons justifying the renewal of fixed-term contracts, has been discussed. Three decisions are important in the context of platform work. In Adeneler the Court found that the concept of ‘objective reasons’ refers ‘to precise and concrete circumstances characterising a given activity, which are therefore capable…of justifying the use of successive fixed-term employment contracts’ ⁷⁷ . The CJEU added that ‘those circumstances may result from the specific nature of the tasks for the performance of which such contracts have been concluded and from the inherent characteristics of those tasks or,…, from pursuit of a legitimate social-policy objective of a Member State’. ⁷⁸

The Court also clarified how the concept of ‘objective reasons’ should not be implemented in national legislation. The question referred to the CJEU related to a national provision foreseeing that the ‘objective reason’ exists if the conclusion of a fixed-term contract is required by a provision of statute or secondary legislation. The Court explained that a provision which merely authorises recourse to successive fixed-term employment contracts in a general and abstract manner by a rule of statute or secondary legislation does not amount to ‘objective reasons’. It added that such a national provision would not be in accordance with the aim of the FTWD to protect workers against employment instability, and render meaningless the principle that contracts of indefinite duration are the general form of employment relationship. ⁷⁹

In Kücük ⁸⁰ the question concerned an employee that worked for the same employer under 13 successive contracts over a period of more than 11 years. The contracts were concluded to replace several permanent employees who were on leave. The Court found that a temporary need for replacement staff may constitute an objective reason. The CJEU stated that: ‘The mere fact that an employer may have to employ temporary replacements on a recurring, or even permanent, basis and that those replacements may also be covered by the hiring of employees under employment contracts of indefinite duration does not mean that there is no objective reason…’. However, the Court added that in the assessment of whether the renewal is justified by an objective reason, the Member State must take account of all the circumstances of the case, including the number and cumulative duration of the fixed-term employment contracts. ⁸¹ Therefore, even if using successive fixed-term contracts to replace permanent employees is, in principle, allowed, the number and total duration of contracts can influence the assessment of the existence of objective reasons.

The practice of the CJEU shows that platform workers as a group cannot be exempted from the protection of Clause 5(1). It derives from Adeneler that the Member States cannot adopt a national provision that would enable to exempt certain arrangements from the protection of Clause 5(1). Hence it cannot be foreseen in national law that an objective justification for the renewal of fixed-term contracts exists if work is performed through platforms. Throughout its rulings the CJEU has stressed the aim of the FTWD to prevent the misuse of successive fixed-term contracts to cover permanent needs and the goal to enable the workers to work in a permanent employment relationship. If the specific nature of the tasks does not prove these to be only temporary, no objective reason for the use of successive contracts exists. Only the replacement of permanent workers has been explicitly regarded as an objective reason in the practice of the CJEU. However, the Court has also stressed the need to take into account the number of renewals as well as the overall duration of the relationship in the assessment.

It is rather clear that in the case of Uber and Deliveroo, the platform does not use the services of the workers to replace permanent staff. These platforms do not have permanent drivers or delivery workers that need replaced; the platform workers form their main workforce. Nevertheless, substituting permanent staff is not the only objective reason; other specifics relating to nature of tasks can also justify the use of successive contracts. If the task is genuinely temporary, successive contracts can be used. It can be argued that the tasks performed through Uber and Deliveroo are, in a sense, temporary, because they last until the clients order the services. Hence, if the user does not request the service, the platform has no work to provide. Often, however, the workers receive requests quite continuously, and the platform appears to have a permanent need for work. Also, as a rule, the provision of work and pay is the main obligation of the employer. The risk of not having any requests cannot be fully transferred to the worker.

Additionally, the CJEU has placed importance on the number and overall duration of fixed-term contracts in assessing whether objective reasons for the use of successive contracts exist. If every task is regarded as a separate fixed-term contract, the number of them can amount to over ten during one day, and several hundred during one month. It is clearly not the intention of the FTWD to allow this kind of series of fixed-term contracts. Yet, the overall duration of these successive contracts may be quite short. If the platform worker uses this option temporarily to balance working and family life, earn money while studying or as a resource for additional income besides permanent work, the worker her/himself can end the relationship in few months’ or years’ time. Would it be justified to use successive fixed-term contracts, for example, if the worker has performed 200 tasks over the period of one month? Although the duration of the whole relationship is not long, according to the CJEU the number of the contracts as well as the overall duration has to be taken into account as a secondary circumstance to assess the existence of objective reasons. The main criterion that needs to be fulfilled is still the temporariness of the work, and this appears to be absent in the case of Uber and food delivery platforms.

The situation, again, can be different in the case of TaskRabbit. For a start, because in this case the user can act as an employer, it is possible that the workers are employed through TaskRabbit to replace permanent staff. The temporary substitution of permanent staff by using successive fixed-term contracts has been allowed by the CJEU. It is even more likely that the same worker works for several users. If she/he has been repeatedly hired on a fixed-term basis by different users, no successive contracts are concluded. Every contract concluded with a different user is a first time fixed-term agreement, the conclusion of which does not need to be objectively justified.

However, in the case of TaskRabbit, too, the use of successive contracts can need additional justification. If the same user hires the worker repeatedly and not for the purposes of temporarily substituting permanent employees, the user needs to find another objective reason to prove that the task is fixed-term. One is, therefore, not allowed to hire the same worker to perform the same task on successive contracts if the factual circumstances prove the user’s need for workforce to be permanent. The fact that the worker has been hired through TaskRabbit does not mean that in this case the abuse of successive contracts should not be prevented. Additionally, the number of successive contracts as well as their overall duration needs to be taken into account. Because of the longer term of separate tasks, the number of successive contracts in the case of TaskRabbit is not as extensive as in the case of the transportation and food delivery platforms. Nevertheless, the overall duration of the relationship can be longer. Hence, even if the same user uses the Tasker two or three times, but cumulatively the relationship lasts for years, we can ask if there is a permanent need for workforce. This situation more likely occurs if the Tasker performs longer-term business tasks under the ‘Jobs’ filter.

If the Member State has chosen to prevent the abuse of successive fixed-term contracts by demanding the existence of objective reasons for their renewal, most often successive short-term contracts in the context of platform work can be avoided. In the case of Uber and Deliveroo, the platform does not have objective reasons to justify such an extensive renewal of extra short-term contracts. In the case of TaskRabbit, the abuse of fixed-term contracts is less likely, but if the same user uses the Tasker repeatedly to perform the same work without replacing permanent staff, the renewal of the contract can be prohibited. This outcome is in accordance with the purpose of the FTWD to enable the workers to be engaged in permanent employment and to treat the use of successive fixed-term contracts as an exemption that should be allowed only if the work is genuinely temporary.

3.2.2 Preventing the abuse of successive contracts in the Member States

Reducing the precariousness of platform workers by preventing the abuse of successive fixed-term contracts is not successful unless the Member States have actually adopted the measures foreseen in the FTWD. The FTWD enables the Member States to choose between three measures to prevent the abuse of successive fixed-term contracts, and they can also combine these measures. The Member States have rather broad autonomy in deciding how the abuse of successive fixed-term contracts can be avoided.

Kamanabrou ⁸² has studied the measures the Member States have adopted. He detected six models used in the Member States. Model 1 combines cumulatively objective reasons and a maximum number of successive contracts, and is in used in France and Luxembourg; Model 2 combines objective reasons with the maximum duration of successive contracts and is adopted in Spain and partly in Sweden; Model 3 restricts the maximum duration of the contracts as well as the maximum number of contracts and is used in the Netherlands and Italy; Model 4 requires only objective reasons and is applied in Austria, Denmark and Finland. Model 5 provides two alternatives: first, the restriction of the total duration as well as the number of contracts, or second, the provision of objective justifications to the renewal; this model is used in Germany, Poland and Belgium. Model 6 limits the maximum duration of the contracts to four years and allows the conclusion of successive contracts after this period only if objective reasons exist; it is used in the United Kingdom (UK), Ireland and Malta. ⁸³

Kamanabrou argues that Models 1, 2, 3 and 4 provide stronger protection against abuse, while Models 5 and 6 allow the conclusion of successive fixed-term contracts more freely. If Model 5 enables the conclusion of fixed-term contracts for a period of two or three years and at the same time places limits on the number of renewals, Model 6 allows the conclusion of fixed-term contracts without justification for up to four years. The level of protection in the case of Model 5 is lower, and in the case of Model 6, considerably lower than in other models. However, the legislative approximation of the Member States is largely achieved because Models 1-5 provide comparable protection. ⁸⁴

The research of Kamanabrou shows that the requirement of objective reasons is mandatorily present in three models and in seven out of 15 Member States that were studied. As discussed earlier, the requirement of objective reasons already quite effectively protects platform workers. Limitations to the maximum number of successive contracts or the overall duration of the relationship enhance the level of protection. For example, in Luxembourg the renewal of the fixed-term contract is allowed twice. ⁸⁵ It would be difficult to imagine the functioning of Uber and Deliveroo if they could use one worker to perform a task only three times. There simply would not be enough workers. As a result, these platforms should opt for concluding permanent employment contracts, which is clearly the aim of the FTWD. It would be possible to use successive fixed-term contracts in the case of TaskRabbit, because the duration of each contract is longer and therefore the amount of the contracts smaller. However, in this case, too, the longest relationships form if the tasks are taken under ‘Jobs’ filter. If the tasks are shorter term, the number of renewals can quickly exceed the limit. Yet, if the tasks are performed for different users, the number of renewals can stay within the permitted range.

If the objective reasons are required and the total duration of the relationship is restricted, platforms have more leeway in concluding successive contracts compared to Model 1. For example, in Spain, objective reasons for the renewal must be provided and the total duration of the contracts cannot exceed 24 months over a period of 30 months. ⁸⁶ In the context of platforms, 24 months is a substantial period of time. In the case of Uber and Deliveroo, an enormous number of fixed-term contracts can be concluded over this period. The maximum duration of the contracts is less protective also in the case of TaskRabbit. Even though in the case of TaskRabbit the tasks are not so similar to those in the transportation and food delivery platforms, they are still short-term: for example, assembling a piece of furniture, helping with moving or cleaning an apartment lasts hours or days rather than months or years. Therefore, the same user can use the Tasker several times to perform the same job before 24 months has passed. Nevertheless, objective reasons for the renewals need to be provided, which makes the conclusion of an extensive number of contracts questionable. It would be questionable whether the same user could use a Tasker to clean her/his house or office every week over two-year period. In this case the user seems to have a need for a permanent workforce. Also, in assessing whether objective reasons exist, according to the case law of the CJEU the number and the overall duration needs to be taken into account.

In some of the Member States studied by Kamanabrou, the maximum number of renewals and the maximum duration of the relationship is defined. For example, in Netherlands the conclusion of three contracts over 24 months is allowed. ⁸⁷ In Germany, the employer can, instead of providing objective reasons, conclude four successive contracts over 24 months. ⁸⁸ In the case of platform work, this option is also sufficiently protective. As already discussed, the number of renewals in platform work is so substantial that it is likely to exceed the number allowed by the legislation of the Member State. In such a situation the employer is obligated to conclude contracts for indefinite period of time.

As pointed out by Kamanabrou, the protection is weakest if only the maximum total duration of successive contracts is restricted, and objective reasons are required at the end of this period. In the UK the maximum total duration of the contracts is four years. During this period an enormous number of short-term contracts can be concluded with platform workers. As no objective justifications for the renewal of contracts is required, the platform in the case of Uber and Deliveroo, or the user in the case of TaskRabbit, can freely use a temporary workforce to complete the same tasks. If that kind of option is chosen to restrict the abuse of successive fixed-term contracts, the aim is most probably not achieved in the case of platform work. Considering that the tasks in platform work are short or very short-term, four years is simply too long a period to for those workers to work in such a precarious situation. However, this option is used only in three Member States out of 15.

To conclude, in most of the Member States the measures taken to prevent the abuse of successive fixed-term contracts are able to fulfil this aim. The options foreseen in Clause 5 (1) of the FTWD effectively prevent the abuse of fixed-term contracts also in the context of platform work. Regarding platform work as fixed-term employment and the application of the measures to prevent the abuse of successive contracts would obligate platforms or users to conclude permanent contracts. This again would substantially improve the labour conditions of platform workers. Nevertheless, in the Member States in which only the maximum total duration of successive contracts is restricted, the aim of the FTWD to prevent the abuse of successive fixed-term contracts is not achieved and the position of platform workers is not improved.

4. Conclusion

Because of the variety of working arrangements in the area of platform work, it is difficult to draw one single conclusion as regards the applicability of the FTWD to platform work and the influence of the measures to prevent the abuse of successive fixed-term contracts on the protection of platform workers.

Yet, in the example of three platforms including Uber providing a transportation service, Deliveroo providing a food delivery service, TaskRabbit mediating different physical services, and AMT as a crowdwork platform, it could be concluded that the FTWD can be applied to some forms of platform work. The FTWD can be applied to the relationship between the platform and the worker in the case of Uber and Deliveroo; in certain conditions the Directive can also be applied to the relationship between the user and the worker in the case of TaskRabbit. In these situations, a bilateral fixed-term employment relationship can be formed between the parties. In the case of AMT, the relationship resembles more of temporary agency work and the TAWD rather than the FTWD should be applied.

The measures provided by the FTWD to prevent the abuse of successive fixed-term contracts effectively protect platform workers if implemented in national law in accordance with the aim of the Directive. The requirement of the objective justification of the renewal of fixed-term contracts accompanied by the interpretation of this measure by the CJEU most likely prohibits the conclusion of very short-term or short-term contracts for each task if the same work is performed. Uber and Deliveroo should, therefore, conclude permanent employment contracts. In the case of TaskRabbit the same user cannot repeatedly conclude fixed-term contracts with the same worker without objective justification.

Also, other measures, such as the maximum number of renewals or the maximum total duration of the successive contracts can effectively protect platform workers if combined with each other or accompanied by the requirement of objective justification. Because of the short term and large number of tasks performed by platform workers the combination of the maximum number of renewals and the requirement of objective justifications is most protective. However, if only the total duration of successive contracts is restricted for certain a period of time (up to four years in some Member States) and the objective justifications are required after this period is exceeded, only a slight improvement of the working conditions of platform workers can be detected. Over the period of one year a platform worker can perform hundreds and thousands tasks and remain in a very precarious situation until the period needing no justification for the renewal of fixed-term contract comes to an end.