COGENS main findings
I. GENERAL FRAMEWORK
A variety of situations
Gig economy is a latent issue, and its regulation is extremely diverse in the Member States of the European Union. This is particularly true for collective bargaining in this field. The variety of legal traditions regarding collective bargaining in the Member States is reflected by the variety of the Member States’ approaches towards (not) regulating collective bargaining for persons working in the gig economy.
Collective bargaining cannot be the single source of regulation in the gig economy, but must be complementary, to a smaller or larger extent, to statutory law. It is clear that constant communication between the different levels of regulation not only at national level, but also between European and national levels is required. The role of public authorities, not necessarily through legislation, is of utmost importance in many countries and new ways are already being explored, in places such as Bologna, Italy, with the development of a local Charter.
When assessing the EU panorama of collective bargaining in the gig economy, the following factors need to be taken into account:
First, gig economy is not developed in the same way in the Member States. Even if it is difficult to identify the real share of persons working under this formula, statistics show that there is a remarkable number of gig-workers in some Member States, such as Italy or Spain. On the other hand, this phenomenon is practically inexistent in other States, for example in Romania or Hungary.
Second, the activities implemented under the name “gig economy” or “platform economy” vary from each other; numerous divergent classifications exist. The clearest classification includes, in one group, so-called “offline activities” that may be reconducted to the idea of “work on demand”, and, in the other group, so-called “online activities” that may be tagged as “crowdwork”. This difference between offline and online activities has direct repercussions on the representation channels for these workers and their coverage under collective bargaining, as depicted by this report. Additionally, online activity entails factors of transnational competence that should not be overlooked when considering future regulations. The Covid-19 pandemic, finally, has blurred in many cases the distinction between telework and crowdwork.
Riders, drivers, and domestic helpers are some of the most characteristic examples of the offline activities. They are usually in the centre of studies on the gig economy, but are also at the centre when considering the collective dimension of labour law in the gig economy. The most remarkable new forms of collective expression can be found in these fields of activities: The first successful experiences of collective bargaining have taken place in these branches of activity. Why? Because of their physical nature, being services that have existed for a long time, and which are now offered on the market under a new model of business, based on platforms and apps. Human proximity leads to the creation of community and that has been the road to a common expression of interests, connecting on-demand-workers with traditional actors: trade unions.
On the other hand, the myriad of activities that can be classified as “crowdwork” are more resistant to the development of the collective dimension of labour law. Collective representation of crowdworkers experiences the same problems previously been detected concerning telework and collective labour relations: Isolation leads to the inexistence of collective voices, even if some practices, such as the “Turker” community may be considered as milestones in the field of representation. In any case, the field of crowdwork and online activities is yet an uncharted territory for collective bargaining.
A third element of variety has its core in the regulation of the employment status of gig economy workers. In the past few years, not only have academics discussed their status of employees or independent workers, but have we seen the emergence of cases brought before courts at national and European level. In some cases, the highest courts were asked to decide, such as in France, Italy, Spain or the United Kingdom. Although the business models of the workers’ business partners varied from each other, the court’s answer was, as a rule, to qualify these workers as employees subject to a contract of employment. Yet, in other cases, the employment status was denied. This was the case in the CJEU’s Yodel-Order; a Belgian ‘tribunal de l’entreprise’ denied the employment status of Uber drivers.
At the moment, the most advanced episode concerning the employment status of gig workers is the Spanish Ley Rider, according to which a legal presumption exists as regards the application of labour law to delivery and distribution activities organized through a digital platform. This has automatic consequences concerning collective bargaining, as these workers are included compulsorily in the respective collective agreements. This Ley Rider is a first small step towards the inclusion of gig economy workers in the general framework of labour regulation under the aegis of statutory law. Much more modestly, and without calling into question the status of these workers, France has recently set up a framework for the organisation of sectoral social dialogue that could eventually lead to possible collective bargaining.
Towards a mixed model of collective bargaining? The personal scope of collective bargaining in the gig economy
The Spanish situation is an exception. In most countries, no explicit piece of legislation exists as regards the legal status of gig economy workers. This has direct implications on the possibility of collective bargaining for gig economy workers. For example, even if since 2016 a part of the French Labour Code is dedicated to workers using a matchmaking platform, they remain a priori independent workers to whom the rules on collective bargaining do not apply.
Traditionally, collective bargaining agreements are concluded to regulate the employment relationship. Thus, on the side of labour, the personal scope is restricted to employees. Independent workers were usually excluded from collective bargaining, except in some countries under specific formulae, such as Germany.
In line with its former case law, the CJEU states in its FNV Kunsten judgment, though, that only those CBA do not infringe EU competition law which are negotiated between „management and labour“. However, the Court also held in FNV Kunsten, that collective bargaining agreements for “service providers in a situation comparable to that of … workers” do not fall within the scope of Article 101 TFEU (para 42). Furthermore, a service provider can lose his status of an undertaking „if he 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” (para 33). This seems to allow for an interpretation in favour of concluding collective bargaining agreements for ‘service providers comparable to workers’ without infringing EU anti-trust law. Yet, uncertainties remain as to which service providers are actually comparable, e.g. Thus, Competition Law is regarded as one of the most serious obstacles on the path towards a new model of collective bargaining with an extended personal scope.
In the light of these uncertainties, the European Commission has recently launched a consultation on the issue of collective bargaining agreements for self-employed. The core question is whether the scope of application of EU competition rules should be insofar limited as collective bargaining agreements concluded for (a yet to be defined group of) solo self-employed should be exempt from EU competition rules. A Communication is expected at the end of 2021 and may bring, or not, a change of paradigm.
To conclude, two main problematic issues can be identified as regards the personal scope of collective bargaining for gig workers. First, the situations in the Member States vary widely from each other. Whereas in many Member Stats, collective bargaining agreements can still be concluded for employees only, in other Member States possibilities exist to conclude collective bargaining agreements also for persons who can be categorized as employee-like. In others, it is possible to declare collective bargaining agreements applicable to non-employees. Other Member States have established models of collective agreements without the same effect as traditional collective bargaining agreements. The overarching problematic, though, is that no uniform definition of those persons for whom – very generally speaking – the scope of collective agreements is extended, exists. This is directly linked to the second main problematic, the interaction between national collective bargaining rules and EU competition law. From a teleological point of view, the CJEU’s case law can be interpreted in favour of exempting collective bargaining agreements concluded for employee-like service providers from EU competition law.
II. SUPRANATIONAL SOURCES
The possible role for EU Law
Whatever the scope of the possible legislation, it is clear that Member States and national social partners are going to be the main actors in this field. However, since the beginning of the COGENS research, a possibility of an EU intervention in the field of gig economy has become more likely and in February 2021 the European Commission launched a consultation on possible action addressing the challenges related to working conditions in platform work.
A hypothetical directive regulating platform work at EU-level could establish that gig economy workers enjoy collective bargaining rights, as the consultation points out, either in an article or a chapter. There is no doubt that the European Union could legislate on collective labour law as mentioned in Article 153.1.f. TFEU. The competence set forth in Article 153.1.b TFEU offers the legal basis necessary to carry out this task if only working conditions are to be regulated, but the mention in Article 153.1.f TFEU to “representation and collective defence of the interests of workers and employers” gives a solid foundation for the development of collective labour law aspects. Of course, this situation will imply the unanimity of the Council to pass a Directive.
This possible development must be, nonetheless, extremely respectful of national competences and traditions. Directive 2002/14/EC offers a clear example for it, establishing a common floor that could be easily implemented in every Member State. The identification of the concrete workers’ representatives must be left to national laws and practices. But the encouragement of specific rules for gig economy workers, such as the “digitalisation” of electoral units or the creation of ways of allocating representation more easily to this type of workers are suitable contents for a supranational regulation.
In any case, the Union could resort to other more audacious avenues. Article 115 TFEU continues to allow, as it has done since 1957 when it was number 100 TEEC, “the approximation of such laws, regulations or administrative provisions of the Member States as directly affect the establishment or functioning of the internal market.” Using this as legal basis, the European Union could approve a directive that creates a minimum level of rights for people working in the platform economy, regardless of their national legal categorization. National competences would thus be respected, rights guaranteed, and there would be a supranational response to a supranational situation.
The human rights perspective: the right to collective bargaining as a human right
The right to bargain collectively is guaranteed as a human right and as a fundamental labour right by many international and European sources, amongst others the ICESCR, several ILO conventions, in particular Nos. 87 and 98, Article 11 of the ECHR or Art 6.2 RESC. With regard to collective bargaining rights for gig economy workers, Art 6.2. RESC is of specific importance: Not only did all EU Member States accept and are bound by Art 6.2 ESC, and therefore have to respect it, but it was the European Committee of Social Rights that held in its recent case Irish Congress of Trade Unions (ICTU) v. Ireland that the decisive criterion as regards the guarantee of collective bargaining rights is ‘rather whether there is an imbalance of power between the providers and engagers of labour’ (…).
It follows from this decision that not only do workers who are employees enjoy the right to collective bargaining, but all providers of labour, including the self-employed who have no substantial influence on the content of contractual conditions. The Committee emphasizes that those persons must be given the possibility of improving the power imbalance through collective bargaining”. In order to clarify, one might add that this is true irrespective of the (formal) status of the provider of labour. In other words, even in those cases where a platform worker is (formally) self-employed, under a teleological and human rights-driven approach she enjoys collective bargaining rights where an imbalance of power between her and the engager of labour exists.
(Old) trade unions and new agents
In most countries, the existing legal framework follows the logic of old patterns of work organization. Thus, implementing collective rights for persons working in the gig economy proves to be rather difficult. A bargaining structure based on the single worksite and the single employer approach, taken together with the majority rule that exists in many Member States, entails structural difficulties in the search for solidarity among platform workers and between platform workers and employees working at the same bargaining unit. This is particularly true for those Member States where bargaining at company level prevails over bargaining at branch level. Without legislative intervention, in many countries, such a decentralised model does not incentivise unions to support platform workers. Discussing platform workers’ rights may be an opportunity to rethink the existing bargaining models, and, in some countries, to encourage industry-level collective bargaining.
In this scenario, after a decade of slow adaptation, old actors seem to be the best placed to integrate traditional prerogatives and new technologies. However, an analysis of reality shows that gig economy workers are sometimes reluctant to join traditional unions. Furthermore, the very unions were, at least at first, not ready to cope with the issue. In order to establish a strong model, the idea of “Smart trade unions” should be brought to debate, which might in the end even result in ‘apps competing against apps’. By promoting debate, raising awareness, creating states of opinion using the very business model of the gig economy, gig workers could be reached more easily. In a model of business based on digital reputations, both for workers and for companies, this part of the activity of trade unions cannot be disregarded.
New agents bring along new ways of collective intervention, too. Yet, these new groups have never succeeded in concluding a collective agreement. They have brought new forms of expression of labour disputes, such as flash mobs, bike-riding demonstrations or blockades. They have indeed created some effects as calls for attention. Nevertheless, they are not regarded as effective bargaining agents.
Branch or company-level agreements
The most probable possibility for as regards the future of bargaining in the gig economy is maintaining the current systems. Their efficacy will greatly depend on the legal classification of platform workers, though. What remains is the dilemma of the bargaining level: Should agreements be concluded at company or at branch level? The answer to this question strongly depends on the strength of the respective negotiating agents.
The panorama of actual collective agreements shows that the first branch agreements concerning gig economy, e.g. the one for hotels and catering in Spain, were concluded by trade unions (and employers’ organizations) that did not take into account the specialities of this activity. They just enlarged their personal scope and absorbed platform economy in the system.
The agreement between the Danish trade union 3F and Dansk Erhverv, the Danish Chamber of Commerce, followed the old pattern concerning actors, too, but went a step further. It was bargained specifically for gig economy workers. At the beginning, only Just Eats’ workers were covered by the agreement, i.e. that it was basically a company level agreement with regard to its personal scope. However, it was then opened to other delivery companies. Thus, it can be considered the first real branch agreement concerning gig economy in its own framework. In Austria, a branch level collective bargaining agreement was concluded for bicycle riders, too, between traditional actors, the chambers of commerce on the employer’s side and the Austrian trade union federation on the other side. Yet, its personal scope is limited to employees only.
The third example is the most remarkable-one: the Italian national-level agreement for delivery of goods carried out by riders. The bargaining agents are a new, specific employers’ organization, AssoDelivery, composed by platforms, and a traditional trade union, UGL, through its specific branch, UGL Rider. The personal scope is also remarkable, as it includes only self-employed riders. This agreement shows the capacity of adaptation and transformation of actors where there is a will to bargain. However, this collective agreement was declared void by a court, considering the lack of capacity of representation of the trade union.
On the other hand, there are some company level agreements, such as the one between the Danish trade union 3F and Hilfr of 2018, or the one in the United Kingdom between GMB Union and Hermes of 2019. The pattern concerning workers’ representation, is the same. The traditional actor is successful in its traditional role when it can put pressure on the other bargaining agents.
Whatever the level of bargaining, clear and direct conclusions can be extracted. In the very few collective agreements that were concluded in the field of gig economy, traditional trade unions were the main actors, following traditional rules.
Finally, the role of employers’ organizations is the opaquest of all. It no longer responds to a class conflict, but of interests, as traditional companies do not share their position and outlook with the new platforms. According to many stakeholders, the position of platforms is not receptive to collective bargaining, but the Italian case shows that there are exceptions to this.
New forms of soft regulation
A curious phenomenon can be observed in several countries, such as Italy, France or Germany, which is the appearance of documents or institutions in the realm of the platform economy that seek to influence it. In none of the cases are they collective bargaining instruments, but because of their originality, in search of a third way that is championed by some, some of them deserve to be examined.
The first in time among these documents was the Bologna Charter, the “Carta dei diritti fondamentali dei lavoratori digitali nel contesto urbano”, of 2018, that had no effective legal value. It consists of twelve articles that are intended to promote safe and dignified employment, but at the same time is compatible with the adaptability of the digital labour market, guaranteeing the improvement of the living and working conditions of service providers. Workers’ representation and labour disputes are present in the Charter, but there is no mention of collective bargaining. In any case, by recognizing the two essential tools for negotiation, the identification of active subjects and the acceptance of pressure measures, the Bologna Charter is creating the right environment for a collective bargaining process to develop as corollary of that recognition.
Germany has provided several examples in this catalogue of related performances, which have an added interest that involves an incursion into the field of crowdworking. First, there is the Code of Conduct “Paid Crowdsourcing for the Better,” signed by several companies, where a unilateral commitment is proclaimed on their part to respect and guarantee a Decalogue of rights. It does not contain any mention of collective bargaining, but its contents approximate in a very important way those that could be subject to it at a more advanced stage. Its most prominent result is the creation of a voluntary conflict resolution mechanism of its own, managed by the IG-Metall union. It only addresses individual conflicts, but to some extent it resembles very prominent outcomes of collective bargaining.
In France, too, the legislator has invited platforms to adopt charters defining working conditions, remuneration for services, and the procedures for information and dialogue with platform workers (delivery personnel or drivers). Although the drafting of these charters is not negotiated or even discussed, the invitation to adopt a more precise framework on working conditions is strong.
Also important is the Frankfurt Paper on Platform-Based Work, signed by seven trade union organizations from Austria, Germany, Denmark, Sweden and the United States, with a very considerable technical team of advisors. Among the essential points that it enumerates is first to respect the relevant collective agreements, but that, in a much more prominent way, it is necessary to insist on the right of workers to organize. A consequence of that right is the ability to negotiate and the declaration that platform operators are appropriate interlocutors to enter into negotiations is particularly relevant.
There is a remarkable consensus among academics and stakeholders that the issue of contents of collective bargaining should be left to the social partners. No legal regulation is deemed necessary in this field.
The analysis of actual contents of collective bargaining in the gig economy provides us with a very diverse answer. The main classification consists of abstract, general and specific contents.
The first group refers to the situation when gig economy workers were incorporated into existing branch agreements, such as the case of Spain concerning hotels & catering. These collective agreements do not contain any specific material rule concerning gig workers, they have just enlarged their personal scope. The analysis of these contents, thus, lacks any interest.
The second group of contents includes traditional categories of regulation that have a special dimension in the gig economy. They include remuneration and working time as the main example: issues that have been always present in the labour market, but present nowadays remarkable characteristics, such as the role of the app in their determination.
The first company-level agreement, signed by the Danish trade union 3F and Hilfr Aps, is the perfect example of this situation, as it covers all the traditional issues. Concerning wages, it contains, for example, the following rule: “Via the platform, the employee can set his/her individual wage. Meanwhile, it can never be lower than the wage stipulated in this collective agreement”. Similar and even more detailed contents can be found in the Italian national-level agreement. They are not innovative clauses, but they are the visible expression of the possible space for collective bargaining agreements in the field of gig economy. In the same agreements, though. there are rules concerning the new, technological aspects of the termination of the employment contract: “Deletion or other depersonalization of the employee’s profile on the platform shall be considered as dismissal”, according to the Danish agreement. Similar contents can also be found in the Italian national-level agreement. It is easy to perceive that this agreement is adapting well-established structures. The majority of platforms, actually, provide the platform with a wide right to suspend the worker or to terminate cooperation, typically without having an obligation to give reasons or under relatively vague criteria only (e.g. referring to rating without stating the acceptable level) and without a notice period. Collective bargaining may be a useful tool in the protection of workers, modulating these wide powers.
A very important field for adaptation is working time. While labour law shall not keep employees and employers from making use of the benefits of modern technologies, the minimum guarantees of working time are necessary for all workers. This does not mean that all traditional legal institutions and limits can apply without adjustment. Nonetheless, adjustment shall not be understood as opt-out: merely technical reasons cannot justify the non-application of working time guarantees. Collective bargaining can be a solution to reconcile the needs of new types of work and working time regulations. Bargaining at collective level can introduce adequate protective measures against an over-emphasis on employer-oriented flexibility and means a more transparent and formal standard setting process than individual negotiations.
As opposed to legislation, the parties of collective bargaining have much better knowledge of the priorities of the effected workplace or sector and the process educes with quick reactions to the rapidly changing market requirements. EU law gives broad space for collective bargaining in setting working time standards. Article 18 of the working time directive allows derogations from the articles on daily rest, breaks, weekly rest period, length of night work and reference periods by means of collective agreements. Empirical evidence shows that working time rules set by collective agreements in digital workplaces is not just an issue for theory. However, it is yet to be seen how parties can make use of the derogation clause in the working time directive and set specific measures designed exactly for gig workers – even within the framework of the employment relationship.
The third group of contents is the more “exploratory” one. It is the field where collective bargaining can be a tool of innovation, dealing with issues that so far have not been a traditional concern. The recent French Décret nº 2021-952, for example, has established a regulation on platform workers’ data and the individual access to them. Workers’ ratings and their control, in a further step, could also be regulated by collective agreements.The negotiation and supervision of the algorithm employed by the platform or the regime of workers’ ratings is the most remarkable of these possibilities.
The Spanish legislation has opened previously a possible way of collective development, as the Ley Rider has included the right of the workers’ representatives “to be informed by the company of the parameters, rules and instructions on which algorithms or artificial intelligence systems are based, that affect decision-making that may have an impact on working conditions, access to and maintenance of employment, including profiling”. It is, obviously, a first, limited step, as it refers to information rights and not collective bargaining. But the gate has been opened for further regulations.
In any case, the right to demand transparency in the decisions and outcomes of AI systems as well as the underlying algorithms must be guaranteed, establishing the right to appeal decisions made by algorithms, and having it reviewed by a human being. Through collective agreements reached by social partners, the parties could address both data input into automated hiring systems and promotion of employees, for example, and employees’ control over the afterlife of the data created by these systems.
Social partners in all sectors could act as spearheads on this issue. In the Green Paper on the Future of Work in Portugal, for example, one of the lines of reflexion is to “encourage, in particular, the regulation of the use of algorithms in the context of collective bargaining, involving the social partners and ensuring the treatment of the matter at the level of collective agreements, in order to ensure an adequate use of AI and to reflect the needs specific to each sector”.
Also, collective agreements could reinforce principles that minimize the new risks associated with the autonomous behaviour of AI, establishing requirements to ensure the protection of privacy and personal data, equality and non-discrimination, ethics, transparency and the explicability of systems based on algorithms, both in terms of the selection of job candidates, and in terms of the execution of the employment contract and the inspection of the worker's professional activity. Furthermore, collective agreements could regulate the consultation of the employees’ unions on AI systems’ implementation, development and deployment.
V. A FINAL CONCLUSION
Our research has shown that actually, collective bargaining exists for workers in the gig economy in the Member States, although differences exist as regards the personal scope of the agreements, the actors involved, the contents regulated and the effects the bargaining agreements have.
Since gig economy and especially crowdworking is as transnational issue, a response from the European Union seems most appropriate. Of course, the Union’s intervention needs to be in accordance with national law and industrial relations practices. Yet, the Treaties offer possibilities to regulate on collective labour law issues.
At national level, traditional actors have managed to conclude collective agreements in the field of gig economy. Although European social partners seem to be reluctant as regards their role of regulating collective bargaining, it seems that trade unions have taken up the challenge of regulating collective bargaining in the gig economy. New actors have not yet lived up to their promises. Our research has shown, too, that agreements were concluded at company level by single employers as well as at branch level mainly by traditional employers’ associations. Yet, in order to create a common floor of rights, branch level agreements seem to be most appropriate.
As regards the contents of collective bargaining agreements in the gig economy, collective bargaining should be used as a tool to provide for detailed regulations. There are specific topics, e.g. the regulation of the applied algorithms, that are most specific to this sector. Thus, a tailor-sized solution agreed upon between management and labour seems to be more appropriate than general rules established by statutory law.
 This paper has been prepared as a synthesis of the COGENS (VS/2019/0084) research project, funded by the European Union. It reflects only the authors' views, and the European Commission is not responsible for any use that may be made of the information it contains. The authors of this paper are José María Miranda Boto and Elisabeth Brameshuber, with contributions from Gábor Kártyás, Barbara Kresal, Teresa Coelho Moreira, Daniel Pérez del Prado and Kübra Doğan Yenisey and materials provided for by the other members of the project’s team.
 Complaint No.123/2016, Decision on the merits of 12 September 2018