Employment relations via the web with international elements: Issues and proposals as to the applicable law and determination of jurisdiction in light of EU rules and principles

1. The issue

The aim of this paper is to explore the question of the applicable law and the competent court within the European Union (EU) in the case of contracts/relationships for work performed via the web and characterised by international elements.

The first thing to note is that the category of web-based work includes not only that now classic type of work resulting from the technological revolution of the 1970s, known as teleworking, in which workers provide their services via the internet and at the same time maintaining a long-term relationship with an employer; but also the new type of work resulting from the emergence of the so-called gig economy, known as crowdworking, in which workers perform via IT platforms activities of short or very short duration for multiple and changing recipients ¹ . The best-known example of crowdwork is Amazon Mechanical Turk, a digital platform based in the USA, which offers its customers services via the internet (e.g. classifying photograph archives, categorising documents, translations etc.), at low cost, since the work is carried out in the form of micro-tasks (gigs) by workers selected by an algorithm (crowdworkers).

Clearly, crowdwork raises many issues for the employment lawyer – including, first and foremost, that of identifying who the actual employer is: the platform or the final recipient of the service. For the purposes of this analysis, the key point relates to the fact that, like teleworkers, crowdworkers can provide their services to counterparts located in other countries, working ‘remotely’, i.e. without the need to physically move from their country of origin. From this it follows that their contract (and relationship) may be characterised by international elements.

On closer inspection this is by no means a new situation, given that working remotely has for some time been possible through older technologies (for example, the case of call centre workers, who perform their work via the telephone). It is, however, clear that the internet, which enables the instant transmission of a huge quantity of data of the most varied nature (written texts, recordings/voice communications, videos, photographs, graphs, numbers, etc.), facilitates and vastly increases the activities that can be performed in this way. This is borne out by the fact that, as many empirical studies report, web-based work is a rapidly growing phenomenon ² .

Hence the importance of the two issues on which the present work will focus.

As is well known, in order to identify the applicable law in the case of an employment contract with international elements, the EU long since identified special connecting factors with respect to the principle of free choice of the parties, which applies to contracts in general. Key among these is the ‘habitual place of work’, which today – thanks to Court of Justice (CJEU) rulings – represents the central criterion for ascertaining both the applicable law (EC Reg. 593/2008, hereinafter Rome I) and the competent court (EC Reg. 1215/2012, hereinafter Brussels I bis).

With regard to web-based workers, the problem that therefore arises is in the first place to understand whether the notion of ‘habitual place of work’ accepted by the legislation on the subject of applicable law and jurisdiction provides adequate answers; or whether, instead, the criterion requires further interpretation through recourse to other rules or principles already present in the system; or, finally, if there remains no alternative but to call for intervention by the EU legislator.

The complexity of the issues under discussion, which relate to distinct, albeit related, juridical sub-systems, led the authors to dedicate separate attention, first to the subject of identifying the applicable law (§§ 2-6), and then to the problem of identifying the competent court (§§ 7-12).

2. A necessary premise: what is meant by ‘individual employment contract’

Before going any further, a clarification is necessary. Since regulations concerning both the applicable law (Rome I, Art. 8), and competent court (Brussels I bis, Art. 20) limit their field of application to the ‘individual employment contract’, the discussion that follows can only hold if it proceeds from the assumption that the single contract (and relationship) taken into consideration can be qualified as such. We must therefore first clarify what is meant by ‘individual employment contract’.

In this regard, the majority interpretation ³ considers that the notion cannot depend on the decisions of individual national systems, at the risk of fragmenting the legal framework, but that it should refer to the broad notion of the employed person developed by the CJEU in applying TFEU Article 45 ⁴ , and based on the performance by the worker of an activity which firstly has an significant economic value, and secondly is subject to some form of directive control.

If we accept this notion of ‘individual employment contract’ as (dependent) employment contract ⁵ , then all contracts in which the directive control of the employer is present fall within the range of application of the two Regulations, regardless of the intensity with which the control is exercised (ongoing, precise and specific orders or broad, general guidelines; instructions related to the content, or to the times and locations, of the work activity); and regardless of the method of explanation and/or transmission (this also includes the directions produced by algorithm and transmitted via the web).

It is therefore important for the interpreter to understand how to legally frame the web-based work contract, which means analysing the specific characteristics of each individual legal relationship taken into consideration. Generally speaking, the problem of definition appears to be more difficult with regard to crowdworkers than for telework: verifying in practice the existence of directive control in fact appears easier when the worker maintains a relationship over time with the counterparty (as happens in teleworking), whereas it appears more difficult when, as a result of the high level of parcelling out of the tasks (typical of crowdwork) the juridical relationship between the parties (regardless of whether it is with the platform or with the subject benefiting from the activity) is of short or very short duration ⁶ .

However, since empirical research shows that very often crowdworkers are subjected by means of algorithms to pervasive powers of management and control by the platform and/or by the recipients of the service ⁷ , we consider that the classification of crowdworkers as employees in the sense affirmed by CJEU case law may in fact be valid in many instances.

For this reason, from here onwards the discussion will be conducted on the assumption that, generally speaking, the contract of web-based workers – teleworkers and crowdworkers – is attributable (via analysis of the specific case) to the broad notion of the ‘individual employment contract’ (as employee) referred to in Rome I and Brussels I bis.

3. The law applicable to web-based workers: the centrality of the ‘habitual place of work’ and its anti-dumping function

We will now analyse the first of the two problems under discussion, that is, what law is applicable in the EU to contracts/relationships of web-based workers who, without moving from their country of origin, work for an employer located in a different country.

As already mentioned, in terms of the applicable law for employment contracts with international elements, the principal reference standard today ⁸ is Rome I, Art. 8 ⁹ , which establishes special connecting factors in this respect, derogating from those generally provided for in contractual matters. The picture is then completed by the subsequent Article 9 of the same Regulation, which defines a corrective mechanism to the operation of Article 8.

It must be pointed out here that the two rules cited above – like the Regulation itself as a whole (see Article 2) – are universal in nature, that is, they are aimed at resolving all conflicts of law, whatever their outcome (applying the rules of a Member State or of a Third State). The interpretation of those rules must, however, take into account two sets of principles and the balancing of those principles: the principle of labour protection, on the one hand, promoted within the EU by Rome I (see § 3); and the principle of free of competition, on the other, which being the heart of the EU Treaties, permeates the system in which Rome I itself operates. Since the ‘weight’ of the two principles – as will become clearer later - is not uniform outside and within the EU ¹⁰ , we will deal separately with the two hypotheses that appear most critical from the point of view of the EU system: that of the worker located in a Third Country who operates via the web for an employer located in an EU Member State (see § 4); and that of the worker located in an EU Member State who performs activities via the web for an employer located in another Member State (see § 5).

First of all, however, we need to examine the rules established by Art. 8 Rome I, and their rationale.

As is known, the provision in question establishes four connecting factors: the first, general and valid for all contracts (see Art. 3 Rome I), is that of the free choice of the parties; in the absence of choice by the parties the secondary criteria are: 1) the ‘habitual place of work’ i.e. the place ‘in which or, failing that, from which the employee habitually carries out his/her work in performance of the employment contract’; 2) the location of the office where the worker is engaged, but only if the ‘habitual place of work’ is not identifiable; 3) the place with respect to which, from the ‘overall circumstances’, it appears that the contract has the ‘closest connection’.

That said, it must be pointed out that on closer examination even in the case in which the parties do in fact avail themselves of the possibility of choice, the secondary criteria mentioned above are nevertheless still of central importance. That same Art. 8 declares that, irrespective of any choice of the parties, the connecting factor can never deprive the worker of the protection guaranteed by the mandatory provisions of the law applicable to the relationship in the absence of choice ¹¹ . Thus, the three secondary criteria mentioned above (i.e. those that identify the applicable law in the absence of choice) become central.

In light of the above, it can readily be seen that ‘habitual place of work’ represents in law the pre-eminent factor, given that it is the first criterion ¹² to apply in order to identify both the applicable law in the absence of choice by the parties, and the mandatory rules that are in any case applicable even where there is choice.

It does not end there, however. The CJEU has intervened over time to reinforce the primacy of the ‘habitual place of work’ (the current text of Art. 8 is in fact the fruit of this case law). ¹³ Indeed, the Court has progressively broadened the meaning of ‘habitual place of work’ to the detriment of the other criteria, first stating that it is not identified with the ‘sole’ place of employment, but with the ‘predominant’ one from a qualitative point of view ¹⁴ – i.e. the place where the worker carries out the ‘essential’ part of his/her duties ¹⁵ – or, alternatively, from a quantitative point of view; ¹⁶ subsequently ruling that it can also be identified with the place from which the worker leaves and to which s/he returns after assignments, in which s/he receives orders and instructions and where the work tools are situated. ¹⁷ Thus expanded, the ‘habitual place of work’ has also become the criterion used in order to identify the applicable law for peripatetic workers, i.e. those workers who constantly operate in different places (such as managers who often move from one country to another; personnel working on board ships; flight attendants and airline pilots).

The broad meaning now attributed to ‘place of work’ inevitably renders marginal the subsidiary criterion of the ‘engaging place of business, ¹⁸ which the CJEU has in fact consistently interpreted in a restrictive way as the place where the stable premises of the company that set up the procedure for conclusion of the contract is located. ¹⁹ Therefore, the formal notion of ‘engaging place of business’ accepted by the Court means that all aspects relating to the management of the work relationship remain extraneous to it (e.g. the place where orders are issued), as they are already subsumed in the criterion of ‘habitual place of work’ (which, as we have stated, is also identified with the place where the worker receives the employer’s instructions). ²⁰

The last criterion stipulated by Art. 8 Rome I, namely the one related to the country with which the contract presents the ‘closest connection’, also appears marginal compared to that of habitual place of work.

Indeed, this criterion, which despite being last on the list is, as will promptly be stated, not least in hierarchy, must be used only with extreme moderation.

The CJEU clarified first of all that Art. 8 places a mere presumption iuris tantum in favour of ‘habitual place of work’ (and secondary to ‘engaging place of business’) as the place that constitutes the ‘centre of gravity’ of the contract, which may therefore be overridden when from the ‘circumstances as a whole’ a closer connection of the contract with another country emerges ²¹ (providing some indications that the judge may use for this). ²² The factor of ‘closest connection’ therefore authorises each of the parties to prove that the contract is ‘more integrated’ in the economic and social context of a different country from the one in which a worker ‘habitually works’ and therefore that this other country is the real ‘centre of gravity of the relationship’.

However, interpretation ²³ has rightly urged judges to use the criterion of closest connection most sparingly: to override the presumption in favour of ‘habitual place of work’ requires rigorous demonstration of the existence of significant, unequivocal and above all ‘non-manipulable’ evidence on the part of the employer that the ‘centre of gravity’ of the relationship is in another country; this could put the choice of applicable law back into the employer’s hands, thus negating all the CJEU case law so far described. It therefore seems reasonable to conclude that the criterion in question can have extremely limited application, most likely only when the parties have a common country of origin. ²⁴

Having summarised the connecting factors, their content and their interrelationships, we will now turn to the question of the overall rationale underlying them. ²⁵

In this regard, the CJEU has repeatedly stated that the criteria are intended to protect the worker as the ‘weaker party’ in the relationship. ²⁶ This is in a certain sense true: they constitute limits to the principle of free choice and therefore prevent the employer from imposing the applicable law on the worker.

At the same time, however, they are not an expression of the principle of favour in a strict sense, because they do not necessarily lead to the application of the law that is most protective for the worker. ²⁷

Rather, the connecting factor of ‘habitual place of work’ expresses a ‘principle of proximity’, because it selects the law within whose system the relationship ‘gravitates’, i.e. the country in which the worker ‘performs his/her economic and social duties’. ²⁸ This ‘principle of proximity’ also lies at the heart of the criterion of ‘closest connection’, ²⁹ which is why the CJEU places it on an equal level with ‘habitual place of work’ (by virtue of the presumptive mechanism already described).

From the worker’s perspective, therefore, the ‘place of work’ is designated to select the ‘closest’, and thus more ‘appropriate’, law (as being the one most familiar and known to him/her).

Instead, from the perspective of the judicial system constituting the ‘centre of gravity’ of the relationship, the ‘place of work’ criterion prevents the deployment of those rules – less protective for the worker (with rules in melius always being accepted) – that the employer would otherwise be able to impose.

A clear anti-dumping function thus emerges from the system, ³⁰ progressively reinforced by CJEU case law through the expansion of the concept of ‘habitual place of work’. ³¹

The parallel restriction of the notion of ‘engaging place of business’, intended to remove from the employer sheer unilateral choice of the law applicable, ³² is consistent with this interpretation, as is the very limited space granted to the criterion of ‘closest connection’.

4. First scenario: the law applicable to web-based workers located in a Third Country. Possible recourse to ‘overriding mandatory provisions’

Having outlined in detail the system’s fundamental components, we will now address the question posed at the beginning: in the EU judicial system, what is the law applicable to employment contracts for web-based workers who operate from a non-EU country?

If, in the absence of contrary provisions, we refer also in this case to Art. 8 Rome I, it is clear that the primary connecting factor of ‘habitual place of work’ always leads to the application of the law of the country in which the worker is physically located. This is in fact – according to CJEU case law – the place ‘in which’ the employee works: where s/he carries out not only the ‘essential’ part but the whole of his/her activity; where s/he receives the employer’s instructions.

Nor does the criterion of ‘closest connection’, which may in some cases determine the application of the law of the common ‘place of origin’ of the parties, help to avoid this outcome.

It follows that with regard to web-based workers, the anti-dumping function underlying Art. 8 Rome I is reduced to zero: ³³ the technology always makes it possible for the employer to choose the least expensive worker, who, all else being equal, will clearly be the one located in the country that guarantees the least protective laws.

This is clearly a totally unsatisfactory outcome, which calls for urgent correction through EU legislative intervention to amend the provisions of Rome I to that end.

However, we believe that even with the legislation as it stands that outcome can be corrected, at least in certain cases, by adopting the line of argument that follows. We are aware that the suggestion can provide only a partial solution to the problem, but it can nonetheless be considered a useful hypothesis not only in and of itself, but also in the (to be hoped for) perspective de iure condendo.

As indicated, the discussion will be conducted focusing first on the case of web-based workers located in a Third Country (in this paragraph) and then on that of web-based workers located in an EU Member State (§ 5).

Which path of interpretation then could be followed in order to preserve the anti-dumping function underlying the criteria laid down in Art. 8 Rome I in the case of web-based workers located outside the EU?

As will be demonstrated in the second part of this article, workers who perform their activities (by whatever means, including via the web) for employers located in the territory of an EU Member State may always choose to petition the jurisdiction of the employer’s domicile (see section II). This is necessarily an expensive option – so the possibility of filing class actions would be very helpful (see § 10) –, but one that the worker, with trade union support, can always exercise.

In this case, Art. 9 Rome I provides a corrective mechanism to the application of the factors referred to in Art. 8: the court must, ‘whatever the law applicable to the relationship’, in any case apply the ‘overriding mandatory provisions’ of the law of jurisdiction, i.e. those ‘provisions the respect for which is regarded as crucial by a country for safeguarding its public interests, such as its political, social or economic organisation’.

There is agreement in the literature on the fact that ‘overriding mandatory provisions’ (or super-imperative norms) constitute a more restricted sub-category than the ‘mandatory provisions’ of the applicable law under Art. 8 Rome I, ³⁴ in that they are distinguished not only by their mandatory nature but also by the fact that they have the precise purpose of preserving the fundamental characteristics of the political, economic and social structure of the country of jurisdiction. Their distinctive character therefore prevents them from including in their scope the entire body of employment law provisions of the country of jurisdiction. ³⁵ Only some of the employment law regulations of the country of jurisdiction fall under the category of ‘overriding mandatory provisions’.

Which ones they are, however, is not identified by the law, which specifies only their fundamental character. Their identification is the responsibility of the judge, who must discern them through reference to all the rules in force, both of the State and of the EU: ³⁶ the rules that derive from the constitutional requirements of the State of jurisdiction and the legal provisions that are their implementation; the rules that derive from the fundamental acts of the Union (Treaties and Charter of Fundamental Rights of the EU) and from the Directives that dictate standards of protection that apply throughout the EU.

It is widely accepted that ‘overriding mandatory provisions’ should include, first and foremost, the rules governing those fundamental rights that safeguard the political and social structure of the State of jurisdiction, such as: the right to strike, the freedom to organise trade unions and collective bargaining, the prohibition of discrimination, the freedom of expression including in the workplace, etc. These are the rights common to the constitutional traditions of the Member States and today solemnly recognised also by the EU Charter of Fundamental Rights.

We share the view of a part of the literature ³⁷ in considering that an essential reference point for identifying the ‘overriding mandatory provisions’ must include ³⁸ the mandatory rules of the State of jurisdiction concerning the subjects listed by the Directive on the matter of the posting of workers (Dir. 96/71/EC, amended by Dir. 2018/957/EU, ³⁹ hereinafter PWD). If it is true that the PWD is not directly applicable to the present case, whether because it is destined to operate only within the EU, or because it concerns only a temporary change of workplace (‘posting’), nonetheless the rules it establishes may be considered indirectly relevant, in that a closer look reveals they fall within the ‘overriding mandatory provisions’ of the State of jurisdiction under Art. 9 Rome I.

Indeed, the PWD complements and corrects the employment law subsystem of private international law precisely in order to preserve the essential features of its anti-dumping function and thus guarantee the fundamental nucleus of the economic and social structure of the host State.

That the PWD complements the employment law system of private international law, confirming the system that underlies it, is evident. Both the PWD and Rome I set forth explicitly ⁴⁰ that in the case of posting, Art. 8 Rome I continues to apply and therefore, in the first place, the principal connecting factor of ‘habitual place of work’, which in the case of posting leads to application of the law of the worker’s country of origin. Posting is by its very nature temporary and therefore does not determine any change in the worker’s ‘habitual place of work’, which therefore is and remains his/her country of origin.

However the PWD immediately supplements and corrects that criterion, requiring that prevalence be given to that part of the mandatory provisions of the host country (i.e. the ‘temporary place of work’, rather than the ‘habitual place of work’) relating to a list of a specific series of subjects, which have a significant impact on the labour costs of posted workers, ⁴¹ with the aim of protecting the internal market from the downward competition of ‘external’ workers recruited at lower cost.

It is clear that the PWD shares the same anti-dumping objective that underpins Art. 8 Rome I.

However, the PWD does not require the application to the posted worker of all the employment law provisions of the country of posting, but only that part – relating to the subjects listed and subject to mandatory provisions effective erga omnes – selected by the EU legislator as being considered essential to preserve the fundamental economic and social structure of the host State. ⁴²

The objective shared by the PWD and Art. 9 Rome I (i.e. to preserve the fundamental characteristics of the economic and social system of the host State) therefore explains why the rules of the country of jurisdiction relating to the subjects listed in the PWD should integrate the ‘overriding mandatory provisions’ under Art. 9 Rome I. The PWD ultimately standardises the irreducible minimum rules in the law of the country of jurisdiction which must, in any case, be applied whatever law is applicable to the relationship.

The reasoning thus far conducted already allows, with the legislation as it stands, web-based workers located in a Third Country who decide to file a claim with an EU court to benefit from the application of a significant body of safeguards specific to the law of the country of jurisdiction, thus guaranteeing the preservation of the fundamental characteristics of the economic and social system, both of the EU as a whole, and of the individual Member States with respect to downward competition from Third Countries. This ‘protective’ outcome, however, does not necessarily penalise the labour market of Third Countries in whose territory the web-based workers are located. Empirical studies have shown, in fact, that web-based work (and work on digital platforms in particular) is currently failing to have an effective positive impact on the economy of developing countries: the sector often employs more highly educated workers, in whom the country’s public system has invested heavily to create a growth engine, but instead ends up side-lining them in repetitive and non-creative tasks, to the impoverishment of their professional skills; remuneration remains relatively low, preventing any positive multiplying effects on the economy of the country of origin ⁴³ and actually increases the economic gap compared to more developed countries.

Ultimately the interpretative path illustrated could have a positive impact across the board, leading to an improvement in the treatment of a section of the workers involved in the global economy (web-based workers located in a Third Country).

We now come to the question of whether the same reasoning can be applied in the consideration of web-based workers located in an EU Member State.

5. Second scenario: the law applicable to web-based workers located in an EU Member State. The need for intervention by the EU legislator

We are aware that the interpretative proposal put forward in the previous paragraph may run into serious obstacles when applied to web-based workers located in an EU Member State, on account of the extremely restrictive scope attributed to the PWD by the CJEU within the internal market and the pre-eminence and, on the other hand, of the principle of freedom of competition compared to the need to protect the individual legal systems of Member States from social dumping.

In fact, as is known, between 2007 and 2008 – in the so-called ‘Laval quartet’ ⁴⁴ – the Court took a favourable stance towards the free provision of services within the internal market, thus restricting the scope of Member States to protect their national labour market. In particular, in the Commission v. Luxembourg ruling the CJEU arrived at the decision that the protection guaranteed by the PWD in the subjects listed therein – to be considered mandatory – constitutes not the minimum, but the maximum protection that can be assured by the host State to the posted worker.

In this interpretation, the PWD is no longer a tool for the protection of the internal labour market of each Member State, but the means that (to some extent) legitimates regulative competition among Member States in order to promote full integration of the single European market.

It is evident that in the face of such an ‘unbalanced’ reading by the Court in favour of economic freedoms it is extremely difficult to imagine that the principles of the PWD may be interpretatively extended, by means of Art. 9 Rome I, to workers within the EU but who fall outside the scope of mandatory application of the Directive, in that they work via the web in their country of origin.

Moreover, it is today the job of the CJEU to interpret not only the PWD, but also the provisions of private international law, which, having been transposed into an EU Regulation, are now EU law.

It is true that in subsequent years the CJEU has somewhat ‘softened’ its position ⁴⁵ and, above all, that the text currently in force of the PWD (as amended by Dir. 2018/957/EU) improves the regulatory framework in the direction of a more effective curb on social dumping. ⁴⁶ It should however be noted that in many respects the solutions offered by the PWD are still a compromise.

It is therefore problematic for the CJEU to extend within the context of the EU the proposed interpretative path of Art. 9 Rome I (see § 4). To achieve that result the CJEU should change its case law orientation.

However, it is not the task of the CJEU but of the EU legislator to determine the balancing point between the two fundamental principles recognised by the Treaty and potentially in conflict with one another – economic freedom on the one hand and employment protection on the other –, establishing accordingly the extent of competition possible among the employment law systems of Member States.

In our view, therefore, web-based work located within the EU territory would require intervention by the EU legislator to establish the minimum levels of protection for workers throughout the EU, irrespective of the place ‘in which’ they physically work (perhaps with specific reference to the subjects detailed by the PWD). In this case, web-based work could represent a first corpus of EU employment law.

6. Preliminary conclusions

The discussion so far conducted with regard to the issue of the law applicable to web-based workers has brought us to diverse conclusions.

In the case of web-based workers located outside the EU who have recourse to an EU judge, a body of mandatory provisions of the competent court may be applied – through the coordinated reading of Art. 9 Rome I and the PWD –, thus according such workers a treatment that is predictably more protective than that of their country of origin, while at the same time protecting the EU system and individual Member States from the downward competition of Third Countries. By contrast, in the case of web-based workers operating within the EU – due to the narrow scope attributed since 2007 by the CJEU to the PWD within the internal EU market – today the law of ‘habitual place of work’ under Art. 8 Rome I must apply, i.e. the law of their country of origin.

We consider the latter outcome unsatisfactory and hope that web-based work can constitute the first segment of a uniform EU-wide employment law (perhaps restricted to an essential core of rules, referring for example to the subjects considered by the PWD), averting the risk of different treatment for web-based workers within the EU and thus of downward competition among the employment law systems of Member States.

Having addressed the question concerning the identification of the law applicable to web-based workers presenting elements of internationality, it now remains to focus on the problem related to the identification of the competent court.

7. The determination of jurisdiction in the Brussels I bis system for transnational employment disputes

In the European context, the determination of the competent court in relation to transnational disputes in the matter of individual employment contracts must be made on the basis of Brussels I bis, which, in section 5, contains some ‘special’ rules, inspired by the principle of protection of the weaker party of the relationship (the worker). ⁴⁷ These rules allow the worker-claimant, who intends to sue the employer domiciled in a Member State, to make use of a plurality of alternative forums. Indeed, according to Art. 21, the connecting factors for identifying the competent courts are: the place of domicile of the employer, or that ‘in which’ or ‘from which’ the worker habitually carries out (or carried out) his/her activity; ⁴⁸ in the alternative, if the employee does not (or did not) habitually carry out his/her work in any one country, the place where the business which engaged the employee is (or was) situated. Finally, in the case of an employer not domiciled in a Member State, the relevant criteria can be the second and the third of the aforementioned ones, or (under Art. 20, para. 2) the place in a Member State where the employer owns a branch, an agency or any other establishment, but only for disputes relating to their exercise. If it is, instead, the employer who intends to sue the worker, Art. 22 gives jurisdiction only to the courts of the Member State in whose territory the worker is domiciled. ⁴⁹ The ‘habitual place of work’ is undoubtedly the most relevant connecting factor used in practice. In fact, the cases in which it is the worker who takes action against the employer are incomparably more frequent than the opposite scenario. ⁵⁰ On the other hand, the CJEU, interpreting the legislative text in the sense of the place not only ‘in which’, but also ‘from which’, the work is carried out, ⁵¹ has, over the years, expanded the notion of ‘habitual place of work’, ending up including the phenomenon of so-called peripatetic work, i.e. the cases in which the worker carries out his/her activity in more than one jurisdiction. The cases of reference are extremely varied: from workers who, although carrying out their activities in more than one State, have a closer connection with only one of them, due, for example, to the existence therein of an office ⁵² or of a place to return to after each trip; ⁵³ to the workers who, regardless of a main centre of interest, carry out their activities in ever-changing places. ⁵⁴

The extension of the scope of application of the ‘habitual place of work’ criterion has entailed a simultaneous compression of that of the ‘place where the business which engaged the employee is or was situated’, considered, moreover, always as residual (and therefore interpreted restrictively). The preference of the CJEU for the ‘habitual place of work’ serves two objectives of primary importance at the same time: on one hand, the protection of the worker; ⁵⁵ on the other hand, the respect for the proximity principle. ⁵⁶

8. ‘Habitual place of work’ in a ‘virtual’ sense: a possible perspective?

The notion of ‘habitual place of work’ has so far been conceived with reference to types of activity mostly characterised by a clearly identifiable physical-material element. ⁵⁷ No references, instead, can be found to wholly or predominantly ‘virtual’ dimensions of the relationship, towards which, due to the impressive technological development of recent years, the world of work is increasingly moving. It is not surprising, therefore, that, with particular reference to the phenomenon of web-based work (and crowdworking in particular), whose characteristics are outlined above (§ 1), the Regulation contains no specific provisions, nor has the CJEU had the opportunity to rule.

One might thus wonder whether the exclusively virtual nature of the working relationship, mirrored in the equally virtual nature of the object of the activity, determines a de-materialisation of the very place ‘in which or from which’ the activity is carried out, possibly to the point of identifying it not so much in the place where the worker is physically operating, ⁵⁸ but in the platform itself, which expands from a mere interactive dimension with the employer – through which to communicate, receive instructions and send information – by virtue of its nature, co-essential and qualifying of the working activity, to the rank of ‘habitual place’ of its implementation.

The need to identify, in the infinite space of the web, an anchorage for the resolution of relevant legal issues, including the identification of the competent court, could however lead to paradoxical results. Since, conventionally, the place of registration of the web platform is deemed the relevant one for detecting its location and its operation, would that place be referred to also for the purpose of identifying the competent court, sub specie of the connecting factor for determining the ‘habitual place of work? ⁵⁹ Indeed, this construction is not entirely without grounds. Mutatis mutandis, in relation to the issue of ‘habitual place of work’ for pilots and flight attendants operating on board aircraft, a minority case law view (also in Italy) suggested identifying it – before the clarifying ruling of the CJEU in the Ryanair case ⁶⁰ – in the place of registration of the aircraft. Although the latter criterion represents a ‘formal’ and easily identifiable anchorage, very often it lacks any close connection with the dispute and the actual location of the itinerant work activity.

However, the reasoning would be unacceptable if applied to the criterion of the ‘habitual place of work’, because it would be fatally at odds with the rationale and the aims of the European legislation, inspired by the principles of protecting the weaker party of the relationship and of proximity. On the other hand, the place of registration would be subject to the risk of manipulation, being able to be placed, at the discretion of the employer, in non-EU countries, characterised by low or even zero levels of worker protection, with serious prejudice to its procedural position. ⁶¹ Finally, adopting this perspective, all the connecting factors would end up overlapping and being interchangeable, leaving the choice of the competent court exclusively in the hands of the employer. In conclusion, the portability of the platform would allow for actual procedural dumping, in addition to favouring phenomena of forum shopping.

On the basis of some recent rulings issued by the CJEU in relation to other situations, but with more than one analogy with those under consideration (harm to individual rights caused through the Internet), ⁶² it certainly appears more reasonable – and in compliance with the fundamental principles of the matter – to opt for the connecting factor in which the centre of main interests of the weaker party of the relationship is located, until eventually making it coincide with the worker’s place of domicile (so-called forum actoris).

In any event, leaving aside for the moment such a wholly ‘virtual’ perspective, at present somewhat controversial, it is appropriate to assess whether, and if so possibly, how, the current European legislation can govern and fit the phenomenon of web-based work in a trans-European dimension, in the dual perspective of identifying the competent judge for cases of web-based workers domiciled in non-EU countries who, however, carry out their activities for an employer domiciled in the EU and, conversely, of web-based workers domiciled in EU countries who carry out their activity for an employer domiciled in countries outside the EU.

9. First scenario: worker via web domiciled in a non-EU country and employer domiciled in the EU

In the first of the scenarios considered, where the claimant is the employer, there is no way to locate judicial proceedings within the EU. Article 22 Brussels I bis is indeed unequivocal in providing, in this case, for the jurisdiction of the courts of the Member State in whose territory the worker is domiciled. ⁶³

Only in the event of a prorogation clause in favour of an EU court would it be possible for the non-EU worker to be sued there: but these are unusual situations in practice and subject to restrictions in favour of the worker, under Art. 23. In the opposite case, where the web-based worker domiciled outside the EU acts against the employer domiciled in the EU, it is likely that the judicial action will be initiated before the courts of his/her own State, which will decide on jurisdiction according to local conflict rules. If the worker decides in any case to bring an action before an EU court, the latter would certainly be competent on the basis of the criterion set out in Art. 21, para. 1, point a) ⁶⁴ and possibly – but it would be a rather infrequent scenario – on the basis of the criterion referred to in Art. 21, para. 1, point b), ⁶⁵ but, in this case, only if recruitment took place at an EU-based office. The jurisdiction should instead be excluded if the worker sued the employer before the courts of a Member State, but not in those of his/her domicile. ⁶⁶ In such a case, the court, in application of the main connecting factor in this matter (‘habitual place of work’), could only decline jurisdiction, whenever the place ‘where or from which’ the worker habitually performs (or has performed) his/her working activity is located outside the EU, ⁶⁷ except in the case the respondent failed to raise the exception of lack of jurisdiction which, pursuant to Art. 26, would definitively site the case before the court seized.

10. Second scenario: web-based worker domiciled in an EU country and employer domiciled in a non-EU country

In the second of the abovementioned scenarios, the action brought by the employer should, under Art. 22, be filed exclusively with the courts of the worker’s domicile. Where the employer acts, instead, before a different court (both in the EU and in a non-EU jurisdiction), net of the admissibility of the action under the local conflict-of-law rules (but only for non-EU jurisdictions), any decision on the merits, made in a State other than that of the worker’s domicile, would almost certainly be refused recognition in any Member State, because it would be in contrast with the special rules on jurisdiction referred to in section 5 (Art. 45, para. 1, point e) i)). In the event that the worker based in Europe intends to bring an action against the employer domiciled in a non-EU country, assuming that the option (although possible) to sue the employer before the court of the latter’s place of domicile is the least favourable and practiced, it will be precisely the criterion of the place ‘in which or from which’ the worker habitually carries out his/her activity that will lead to the initiation of judicial proceedings before an EU judge, most likely that of the worker’s domicile. ⁶⁸

11. The role of trade associations and collective redress

In all cases where the operation of the different connecting factors previously examined required the worker to bring an action in a jurisdiction other than that of his/her domicile, his/her procedural position (and, consequently, the protection of his/her rights) may be affected, since s/he could no longer avail her/himself of the nearest forum. Initiating and dealing with litigation abroad, in fact, often requires not only financial but also informative tools, of which the worker-claimant ⁶⁹ who acts individually is likely to be lacking. The contractual weakness of workers operating on, and through, the web, in particular, appears even more marked, because, unlike those of a particular company, they rarely know each other, hardly ever coming into contact personally. This affects the possibility of creating bonds and relationships of trust, which are essential for coordinating joint actions to protect the category.

In relation to disputes in which they are involved, therefore, perhaps more than in other cases, a role of primary importance should be played by trade associations (first and foremost, the trade unions), which can more effectively handle cross-border litigation (see the case of posted workers), ⁷⁰ also and especially in the context of class actions. ⁷¹

Net of the differences in the respective legal regime, what characterises this mechanism is the fact of allowing a group of subjects (‘class’), harmed by contractual and/or extra-contractual violations, to bring ⁷² against the author of the harmful conduct a single judicial proceeding, which will involve, on behalf of the class, the participation of a single subject (the class representative). This mechanism makes it possible to overcome certain barriers to access to justice and the proliferation of parallel individual proceedings, originating from the same factual circumstances. ⁷³ Within a collective redress, the effects of the decision will be felt by all those who have (spontaneously) joined the action. ⁷⁴ Finally, this mechanism represents an effective deterrent for companies (incentivised to internalise costs and avoid negligent and/or fraudulent conduct) and, through the containment of administration costs due to economies of scale, an instrument that increases efficiency of the system as a whole. ⁷⁵

The Italian legal system has recently regulated the phenomenon with a specific law, ⁷⁶ not without weaknesses, ⁷⁷ which does not however contain, as far as it is of interest here, any provision on possible cross-border profiles of collective redress. ⁷⁸ This is a shortcoming common to many European legal systems, ⁷⁹ which is in stark contrast with the exponential increase in risk, caused by economic globalisation and increasing cross-border trade, that collective interests located in more than one Member State be could affected by violations of EU law. ⁸⁰

Moreover, at European level, despite the remarkable initiative of the EU institutions, to date there is no binding legislative instrument on class actions (or collective redress). ⁸¹ Among the EU-related initiatives worth mentioning is, in particular, the Recommendation issued on 11 June 2013, ⁸² with which the Commission outlined a set of common principles for collective redress mechanisms, with a view to harmonising national procedures and coordinating them in respect of cross-border disputes.

A series of recent Reports and Studies, ⁸³ however, show starkly that the implementation of the provisions of the 2013 Recommendation by Member States has not been at all satisfactory and the current situation is anything but harmonised. The national disciplines diverge, even very strongly, with reference to practically all the procedural profiles of the collective redress mechanisms, ⁸⁴

On the other hand, as far as the determination of jurisdiction is concerned, national legislations, which are deficient in this respect, merely refer to Brussels I bis, which, however, having essentially been designed for individual litigation only, ⁹¹ does not contain any special jurisdictional rule for collective redress mechanisms. ⁹² The application of its rules, which can be extended to the cases we are dealing with here only by analogy, ⁹³ makes the bringing of a cross-border collective judicial action extremely problematic, if not impossible. ⁹⁴ In assessing the admissibility of such actions, in deciding on all procedural issues that may arise during the proceedings, ⁹⁵ or in deciding whether or not to recognise decisions issued in another Member State at the end of class action proceedings, in fact, courts will have to assess the existence of the requirements in relation to the individual parties, to the single petita and to the single causae petendi, with potentially disruptive effects of the collective actions, frustrating their rationale and goals.

The difficulties that exist in a European context – where integration of legal systems and cross-border cooperation are particularly intense – are multiplied when considering the international (extra-EU) dimension, ⁹⁶ which is inevitably subject to the conflict-of-law rules of individual States. ⁹⁷ The outcomes relating to the different procedural steps of a collective judicial action are in this case unforeseeable, fatally variable from State to State and subject to potential discrimination between subjects belonging to the same class. ⁹⁸ The situation could perhaps improve with the Hague Convention on the Recognition and Enforcement of Foreign Judgments (adopted on 2 July 2019), which includes, in the notion of ‘decision on the merits’, also decisions issued within class action proceedings, ⁹⁹ allowing their easier recognition and enforcement in other Contracting States. A high number of ratifications, ¹⁰⁰ in fact, could have an indirectly harmonising effect on the main procedural profiles of the collective redress mechanisms, with a gain in terms of legal certainty.

However, at present, the actual impact of collective judicial actions on the cross-border protection of web-based workers’ rights is far from ideal. And this, as we have seen, concerns both European and, even more so, non-European workers.

12. Second conclusions

The foregoing analysis should have highlighted some critical issues in respect of the identification of the competent forum in relation to the phenomenon of web-based work in general, and of crowdworking in particular, which is not yet sufficiently regulated, in particular as far as the international procedural dimension is concerned. Indeed, the extension to those new scenarios of common rules and principles developed with reference to other contractual typologies (primarily those of so-called peripatetic workers) is not always satisfactory, due to the extreme variety of disputes in which workers of this category may be involved.

Moreover, due to the barriers to access to justice that the worker is faced with (especially in relation to an individual dispute to be located in a jurisdiction other than that of his/her domicile), an attempt was made, in section II of this article, to enhance the role of trade unions as a collective pursuer of judicial remedies. In respect of the latter aspect, dissatisfaction with the current regulatory regime regarding class actions, when their use is necessary in the context of cross-border disputes, should impose an acceleration in the path of harmonisation of national disciplines ¹⁰¹ that, for example, foresees, for an entity or association registered in a particular State, the possibility of representing applicants from different States (even outside the EU) and of operating in any jurisdiction, with the possibility that other associations intervene in the same proceedings.

In order to avoid the phenomenon of parallelism of procedures, one could even envisage a system in which, in the face of prejudice suffered by a plurality of workers, only one collective action is authorised at European level (possibly subject to a single law), to be filed not so much with the court of the respondent’s domicile (which tends to favour the respondent to the detriment of the claimants), but before the court of the place where, for reasons closely related to the type of harmful conduct or to that particular class of applicants, the main centre of their interests is to be located. ¹⁰² This should be combined with intense information activity (through publicly accessible and interconnected electronic registers), aimed at making all potential stakeholders aware of the scope of the action (or actions). ¹⁰³

These, we are well aware, are almost futuristic solutions at the moment, as, in the European context, they are up against the predictable resistance of the Member States as well as the principles of proportionality and subsidiarity. Moreover, in relations with Third Countries, a critical issue is the absence of international agreements of mutual assistance and judicial cooperation (since the Hague Convention on the Recognition and Enforcement of Foreign Judgments is inconclusive for these purposes). But this, we believe, is definitely the ideal direction in which we should move.