- Gionata Golo Cavallini
- Doutorando em direito de trabalho pela Universidade de Milão
- Graduado em direito pela Universidade de Milão
Artigo recebido em 25/5/2017 – Aprovado em 30/5/2017
RESUMO: O contributo mira a examinar os perfis jurídicos do “trabalho digital”, inclusivo do trabalho “on-demand-via-app” e do “crowdwork”, novas tipologias de trabalho na “gig-economy” que envolvem, respetivamente, uma mão-de-obra on-demand móbil e uma mão-de-obra on-demand virtual. O surgimento do “trabalho digital” e a “commodificação do trabalho” envolvida afetam não apenas as regras e os princípios do direito do trabalho “tradicional” mas até o conceito fundamental que o trabalho não é uma mercadoria.
Não obstante disso, o “desafio” do “trabalho digital” pode também constituir uma ocasião para o desenvolvimento, mesmo a um nível interpretativo e de policy making, de esquemas protetivos aplicáveis a todos os seres humanos que trabalham, independentemente da natura da relação jurídica em virtude da qual desempenham as próprias atividades. Nessa perspetiva, o contributo quer verificar se é possível e útil pesquisar disposições protetivas fora do âmbito do direito do trabalho subordinado. A tal fim, os recentes desenvolvimentos da jurisprudência e da doutrina da área anglo-americana (nomeadamente, os casos relativos a Uber) serão comparados com a experiência, mais restritiva, dos tribunais italianos em matéria de qualificação da relação de emprego.
O contributo vai depois pesquisar os possíveis remédios protetivos previstos pela disciplina geral dos contratos, pelas regulações do trabalho autónomo e pela disciplina dos contratos entre empresas caraterizadas pelo deslocamento de poder contratual. À luz dessas disciplinas alguns pontos críticos serão examinados, como a legitimidade da desativação do account do trabalhador e da clausula que permite recusar o trabalho prestado pelo “crowdworker”, sem compenso, para verificar se os remédios extra-laborais podem procurar alguma proteção, e, no caso afirmativo, até qual extensão.
PALAVRAS-CHAVE: Novas tipologias de trabalho. Trabalho digital. Trabalho autónomo. Trabalho subordinado. Proteção trabalhista
ABSTRACT: The paper aims to analyse the legal aspects of “digital work”, as including “work on demand via app” and “crowdwork”, involving, respectively, an “on-demand mobile workforce” and an “on-demand virtual workforce”. The breakout of digital work and the “commodification of labour” jeopardise, together with labour law standards, also the truly founding idea that labour is not a commodity.
Nonetheless, the “challenge” of digital work may constitute an opportunity for further reflections on the development of protective schemes applicable to all human beings who work, regardless of the legal scheme under which they carry out their activities. To verify whether it is useful and possible to search for some protective provisions for digital workers outside statutory employment law, the paper will compare the recent developments of Anglo-American literature and case law with the less open-minded perspective of Italian Law.
The paper will therefore propose try to find out whether it is possible to elaborate some protective schemes for digital workers through the recourse to contract law and self-employed work regulations. Under such regulations, the paper will deal with some critical points (such as the possibility to refuse Amazon Mechanical Turkers’ work without payment, or to “dismiss” an Uber driver for low reputational rates), in order to verify whether outside labour law it is possible to find some protection for digital workers, and, in the affirmative, to what extent.
KEY-WORDS: New type of labour relationships. Digital work. Autonomous work. Subordinate work. Labour protection.
Summary: Foreword: the qualification of the labour relationship as a presupposition for the individuation of the applicable social security provisions. –1. Introduction: the “nosedive” of labour law? – 2. Qualification Issues. – 2.1. Are digital workers employees under Italian law? – 2.1.2. From ponies to kangaroos. “Work on demand via app” in the light of the Italian case law on pony expresses. – 2.1.2. When your cubicle is at home. “Crowdwork” in the light of the Italian decisions on call center workers. – 2.2. The unsatisfactory recourse to intermediate categories. – 2.3. Self-employed workers in many cases, unfortunately. 2.4. Mere intermediary or party to the self-employment contract? – 3. Self-employed work rights and their sources. – 3.1. Some consequences under private law. – 3.2. Towards the development of a Statute for self-employed (digital) workers. – 4. Conclusions.
Foreword: the qualification of the labour relationship as a presupposition for the individuation of the applicable social security provisions
The present paper deals with the qualification of the new type of labour relationships emerging in the so-called gig economy (or platform economy), a scenario dominated by platforms like Uber, Taskrabbit, Deliveroo, AMT. Such companies platforms developed a business model that challenges the traditional perspective of labour law, as it introduces elements of rupture with the received notions of employer and employee. Platforms do retain an extremely significant degree on control over the “digital workers” performances, but, at the same time, the worker enjoys a degree of spatial and/or temporal flexibility that does not fit so well with the notion of subordination as developed in the last century.
In some jurisdictions labour judges tried to prevent “digital workers” form falling into an “empty space of law” by re-qualifying their relationships with the platforms as employment relationships. The litigation related to the status of Uber drivers reached even Brazil, with a Minas Gerais first instance judge according requalification, and a second instance court overruling the first decision.
The problem of the qualification of the relationship as a subordinate employment relationship or as a provision of self-employed work is an evergreen issue for labour lawyers in every jurisdiction, and it does indeed have extremely relevant consequences with respect to the application of different social security regimes.
Therefore, not only the qualification of the relationship has an impact on the degree of substantial protection that law grants to the worker, but also on the juridical position of the worker before social security authorities. Employees do generally enjoy several benefits (with respect to the payment of social contributions, which is normally set on the employer, as well as to unemployment benefits) that autonomous workers do not, on the ground of their supposed higher independence in the market.
Qualification issues, which traditionally belong to the domain of “pure” labour law, are therefore crucial indeed also for social security law. In this perspective, although the reflections developed in this paper do not deal directly with social security issues, they still represent an attempt to construe a preliminary basis that may be useful also for further investigations in the field of social security law.
1. Introduction: the “nosedive” of labour law?
Nosedive, the first episode of the third season of the British series Black Mirror, released worldwide on Netflix in October 2016, depicts a dystopian reality where everyone can give a score to anyone else through a five-star system implemented on their smartphones, displaying everyone’s name and current rating. As personal rating determines social status and access to jobs and housing, Lacie spends her day handling frenetically her smartphone trying to improve her rating, until she goes through a sudden and unexpected rating decrease. Such “nosedive” will drive her to madness while the spectator falls as well into an increasing anxiety towards a sinister and yet believable reality. The episode witnesses quite precisely how “digitalisation” is not just the breakout of new organisational and productive schemes. It interferes with the very essence of human life, reshaping the invisible borders between work time and free time, work place and home, and whoever gets work mails on his/her smartphone, at any time of the day and of the night, may confirm.
Lightening the risks of reputational systems, Nosedive makes a fine and yet clear reference to Uber’s five-star rating system, one of the core points of the organisational schemes of the famous American Platform. Reputational systems constitute a leitmotiv of most of the platforms that provide services that can fall within the notion of “digital work”, as including “crowdwork” and “work on demand via apps”, the two main categories that have been identified in literature as part of a unitary phenomenon, calling for a unitary approach.
The similarities between the legal functioning of different platforms offering “digital work” services suggest indeed to adopt a same perspective in order to adequately take on the very same challenge brought by both types of “digital work”: the challenge to avoid technological (r)evolution from bringing also labour standards to a “nosedive”, allowing the rise of unregulated legal schemes apparently able to bypass statutory employment law by introducing elements of rupture with the traditional notions of employer and employee.
In fact, platforms claim they operate under a legal scheme where, at a first glance, it looks like there is no room for the application of labour law. Such scheme, notwithstanding the several differences among the various work-on-demand platforms, could be in most cases depicted as follows:
In the perspective of avoiding this structure – which is, at least, theoretically conceivable – from entailing the breakout of a model where labour seems to fall into an “empty space of law”, the paper aims to verify whether it is possible and useful to search for some protective provisions applicable to digital workers outside the field of statutory employment law.
The path of the reclassification, with the consequent application of statutory employment law, would be of course the most logical strategy in order to provide adequate protection to platform-mediated workers. This is the perspective of the first European decision on the status of Uber drivers, issued by London’s Employment Tribunal on the 28th of October 2016 (exactly one week later after the release of Nosedive). Emphasising inter alia Uber’s reputational system, Judge Snelson ruled that the relationships between the platform and the drivers are subject to statutory employment law provisions on minimum wage and paid leave.
However, the challenge of “digitalisation”, in its tending towards a dangerous commodification of labour, requires action on different frontlines, as the strategy of reclassification cannot be always adequate, correct or sufficient. Even Judge Snelson admitted that Uber “could have devised a business model not involving [it] employing drivers”. Moreover, we are addressing a global phaenomenon, which has also to pass through the lenses of different legal systems, characterised by relevant differences with concern to the criteria employment judges would use to qualify the relationship.
Thus, if one part of the strategy is definitely to return – where possible – to the domain of statutory employment law those relationships that are actually misclassified under applicable legislation, it seems necessary to conceive also a complementary path, especially in those jurisdictions where judicial reclassification would be difficult to reach.
This happens to be the case of Italy, where first instance judges tend to give extreme importance to the circumstance that the worker is not technically bound to perform his/her tasks, supported by the Supreme Court enduring statement that “any human activity can be performed under the scheme of an employment relationship or under the scheme of self-employed work”. In such jurisdiction, therefore, it seems necessary to make further reflections on the development – both at an interpretative and at a policy making level – of protective schemes applicable to all human beings that work, regardless of the legal scheme (employment, self-employment or other) under which they carry out their activities.
To this end, the first part of the paper (meant as a pars destruens) will deal with the problem of the qualification of the relationships involving the worker, the user and the platform, and will conclude that – at least under Italian law – many platforms may successfully claim the self-employed status of their workers. Even the existence of intermediation relationships seems convincing, although the platform does not carry only intermediary’s obligations.
The second part of the paper (in the attempt to be a pars construens)will then analyse the consequences of such reconstruction under Italian law, with particular reference to some critical points (such as the possibility to “dismiss” an Uber driver because of his/her low reputational rates or to refuse Turkers’ work without payment). The paper will try to give an answer to those open questions through the application of general contract law, self-employed work rules and B2b regulations, with an eye to two recent Italian legislative proposals, in order to verify whether some protection can be found outside the domain of statutory employment law, and, in the affirmative, to what extent.
2. Qualification Issues
The qualification of the relationship as an employment or self-employment one represents a crucial standpoint in almost every jurisdiction. Employees generally enjoy several statutory provisions (on wages, working time, and social security benefits) that independent contractors do not, on the ground of their supposed higher bargaining power and economic independence.
It is quite difficult to qualify digital workers univocally either as employees or as independent contractors, as they find themselves in some sort of grey area. The platform operates at the same time as a broker matching labour supply and demand, as a provider of services and goods and as an employer establishing the most important rules governing the transaction, including its termination, which may consist in the deactivation of the worker’s account.
As one of the first decisions from the U.S. litigation on platform drivers pointed out very clearly, “Lyft drivers don’t seem much like employees […] but Lyft drivers don’t seem much like independent contractors either”. “A reasonable jury could conclude that the plaintiff Lyft drivers were employees. But […] a reasonable jury could also conclude that they were independent contractors”.
In the U.S., the difficulties in reaching a clear consensus on the legal status of digital workers brought to significant litigation, which appeared to undermine the entrepreneurial model adopted by the “work on demand via app” platforms. Notwithstanding the worldwide debate it gave raise to, as many platforms accepted to negotiate a millionaire settlement (or even unilaterally decided to acknowledge the employee status of their workers), the question is still open.
The recent judgement issued by London’s Employment Tribunal took a clear position stating the applicability of the statutory rights related to minimum wage and paid leave to Uber drivers working in the London area, individuating a series of circumstances pointing to their “workers” status. The anti-formalistic approach of the British judge in the case reflects the efforts of that literature that suggested to determine the scope of statutory employment law adopting a functional approach to the concept of Employer, instead of recurring to the “received” notions of employee, even with particular reference to the case of platform work.
2.1. Are digital workers employees under Italian law?
In Italy, gig-economy-related litigation raised first on competition law issues, with licensed taxi drivers successfully preventing Uber from releasing the Uber-pop service, which would have allowed (unlicensed) private citizens to provide transportation services. The Italian gig-economy faced its first defy from the labour side just in October 2016, when Foodora drivers took collective action in reply to the decision of the Company to change their payment scheme from a gross 5,60 Euros per hour to a gross 3 Euros per delivery. The mobilisation of Foodora drivers brought the problem of digital labour to the centre of the debate, gaining also the cover of the prestigious weekly magazine Internazionale, and required the intervention of the Labour Department, after which the company increased the delivery fee to four gross Euros.
Even though in Italy the ascertainment of the employment status follows a path which is similar to the several tests developed in Common Law systems, as it focuses on the degree of control that the employer exercises on the execution of the performance, to seek the re-classification of many digital workers as employees under Italian law could be somewhat “gasping”.
While common law jurisdictions developed a series of different “tests” to determine the application of the different statutory regulations applicable to the employment relationship, Italian law, as many other continental laws, provides for a unitary notion of employee, i.e. “who engaged himself to cooperate for remuneration in an enterprise by working manually or intellectually under the direction of the entrepreneur”. The identification of the characters of the employee’s subordination – as opposite to the self-employed worker’s autonomy – has always been an evergreen topic, accompanying the development of Italian labour law from its very beginning to the challenges brought by technological innovation.
The reasoning adopted by case law and administrative authorities on the qualification of the status of pony expresses and of call center workers represented an important step in the elaboration of the criteria used to identify employment relationships. Today, the reasoning developed in those cases looks like the most persuasive argument for whoever would have to defend before a court the self-employment status of digital workers.
2.1.1. From ponies to kangaroos. “Work on demand via app” in the light of the Italian case law on pony expresses
If you just change his walky-talky with a smartphone, a pony express from the roaring 80s presents many similarities with those forms of “work on demand via apps” that provide delivering services (such as Deliveroo, Foodora and JustEat, whose drivers you can easily spot in many Italian streets).
A significant litigation accompanied the development of the pony express business model, together with an animated doctrinal debate. Several first instance judges acknowledged the existence of employment relationships, in labour proceedings promoted by the worker or by the Social Security Authority (INPS)  as well as in criminal trials. Such decisions represented the attempt to interpret the legal notion of employee as an open reference “to the economic and social reality in its variety and historical dynamicity”. Emphasising the economic dependency of the worker, his/her insertion into an entrepreneurial organisation, the degree of control exercised by the company and the continuity of the performance, those decisions deemed irrelevant the allegation that workers were free to accept or refuse the single tasks assigned, rejecting the companies’ main defence. The reasoning made more than thirty years ago that “it is not realistic to sustain that messengers are free to accept or decline the single task. […] once he chooses to work to gain money, the messenger is actually bound to answer the call to perform the delivery”, resembles some of the considerations made by Common Law judges in the Uber proceedings.
Higher courts, however, did not embrace this interpretative option and overruled the first instance decisions, individuating precisely in the freedom to refuse tasks the main element excluding the existence of a tie of subordination. The fact that in many cases the pony express carried out the activity continuously was deemed irrelevant, supported by those Authors pointing out that the employment status finds its reason in a legal-only notion of continuity, i.e. in the “legitimate expectation of the creditor on the continuity of the performance according to a program agreed in advance”.
Since higher courts continue to uphold this orientation, it seems very difficult for the many digital workers occupied in the sector of transportation and delivering services to achieve the judicial ascertainment of the employee status under Italian law.
2.1.2. When your cubicle is at home. “Crowdwork” in the light of the Italian decisions on call center workers
Call centers were at the centre of literature’s reflections in the first decade of this century. As they gave rise to the most massive concentration of allegedly self-employed workers in a particular economic sector, call centers became emblematic of the condition of precariousness of many Italian workers, deserving also a role in popular culture and cinema.
Some of the outcomes reached by legal literature, case law and public authorities should be taken into consideration when it comes to the qualification of crowdworkers, i.e. those digital workers who do not only meet their tasks but also perform them online, constituting a global virtual workforce.
In fact, there is not that much difference between the human intelligence tasks crowdsourced through Amazon Mechanical Turk and other crowdsourcing platforms and some of the tasks performed in call centers. In both cases, we are mostly before labour intensive activities involving the execution of monotone and repetitive “microtasks” that do not require particular skills.
Also due to the political relevance of the problem of repressing misguided employment relationships, during the centre-left Prodi Government (2006-2008) and the centre-right Berlusconi government (2008-2011) the Italian Ministry of Labour repeatedly issued interpretative criteria to determine the conditions under which it is possible to work in a call center under a self-employment relationship.
A 2006 circular addressed to labour inspectors clarified that only call center workers who perform in bound activities – i.e. who undertake to answer to incoming calls – shall be always deemed as employees. With reference to out bound workers – i.e. those who undertake a campaigning project consisting in making a certain amount of calls – the circular stated that it is possible to qualify the relationship as a self-employed one insofar as the worker is free “a) to decide whether to perform the activity and when; b) to schedule the daily working time; c) to suspend the execution of the performance”. In this case, safe for forms of coordination with the client, the out bound worker can determine autonomously his working schedule and therefore falls beyond the scope of the employee notion. Although the Ministry is not a Legislator and was just addressing labour inspectors, the document had a significant impact also on case law, with some decisions deeming as self-employed out bound workers and other decisions emphasising the non-binding nature of the ministerial document.
Aware of the difficulties in applying the criterion based on the distinction between in boundand out bound activities, the Ministry issued in 2008 a second circular which narrowed the scope of self-employed work in call centers, individuating a series of factors which would entail reclassification of out bound self-employed workers. The successive centre-right government, however, clarified that such sort of presumption of the existence of an employment relationship contrasted with the discipline of self-employed project-related work provided for by legislative decree 276/2003, as well as with the Supreme Court jurisprudence.
Even adopting the broadest approach promoted by the first document, however, the subjection to the direction of the employer also with regard to the definition of the working period still represents an element that is necessary to claim the employment status successfully.
In light of these principles, it would be even more difficult in most cases to classify as employees, under Italian law, those workers who perform their activity on crowdsourcing platforms. Not only they are free – like the out bound call center workers – to determine their working schedule in terms of time, but they also retain “the freedom to choose when and where to work, how long to spend, and what work to perform”. In addition, the fact that they perform their activity from their own homes, or from any place where a Wi-Fi connection is available – thus without any physical relationship with any workplace – would constitute a further element that an Italian judge may valorise in order to deny reclassification.
2.2. The unsatisfactory recourse to intermediate categories
Pony expresses and call center workers case law shows that – despite the attempts to valorise the economic and social weakness of the worker – Italian labour law developed a legal-only notion of subordination, meant as the provision of a personal effort, in terms of time and energies, to the employer and subject to his direction. Even though judges would evaluate the circumstance of the effective and stable introduction of the worker in the firm’s organisation and a series of secondary criteria, the autonomy of the worker with respect to time scheduling, choice of tasks and working place appears to be de iure condito an unsurmountable obstacle to reclassification in terms of an employment relationship.
It is also hard to assimilate properly digital workers to the category of “quasi-subordinate workers” or “dependent contractors” developed in Italy, German and Spain, as some scholars suggested. In general, the notion of “economic dependency” postulates that the worker devotes the main part of his activity to a single client, while in the case of digital work, as it has been noted, there is often no stable counterparty to burden with duties and responsibilities.
The scope of the Italian definition of quasi-subordinate workers includes those workers who, without any tie of subordination, provide continuously a mostly personal activity under the coordination of the counterpart of the contract (art. 409 civil procedure code, as emended in 1973 for the purpose of the extension of a few employment warranties to quasi-subordinate workers). Even though coordinated and continuous collaborations were treated as self-employed relationships, some particular rules were set with regard to social security contributions, which are carried for 2/3 by the client (art. 2 (30) Law 335/1995). The so-called Biagi Reform (Legislative Decree 276/2003) provided for a series of measures in the attempt to preserve the autonomous character of such collaboration agreements, which had to be related to a specific “project”, under penalty of reclassification.
The recent reform of Italian labour law known as Jobs Act, however, abrogated the whole discipline on project-related work (art. 61-69.bis, Legislative Decree 276/2003) that was applicable to quasi-subordinate workers. Such abrogation, together with the parallel re-conduction to the field of employment of so-called hetero-organised relationships, has been waved by the government as the elimination of a precarious and unpopular non-standard contract .
However, it is to say that, on one hand, the last measure may not in fact contain any real innovation, as courts already used the criterion of temporal and spatial hetero-organisation. On the other hand, what has been eliminated is not the possibility to recur to quasi-subordinate work but just the few warranties that had been introduced to avoid its abusive recourse (such as the duty to indicate the specific “project” for which the contract is stipulated).
If we consider the counterpart of the self-employed digital worker to be the several clients that he may happen to serve, it would be quite difficult to conclude that such activity, which is certainly personal, is characterised by the elements of continuity and coordination. The terms and conditions of some platforms appear to foresee the risks for a single client to repeatedly receive services from a same worker, and decline any responsibility for the case that such continuous recourse entails the constitution of an employment relationship under applicable legislation.
If we evaluate the existence of the elements of continuity and coordination with reference to the relationship between the worker and the platform, instead, we could easily conclude that in many cases there is a quasi-subordinate relationship falling within the scope of art. 409, n. 3, of the Civil Procedure Code. Yet such classification would not provide the worker (rectius, the collaborator) with any significant degree of substantial protection.
In Italy, some platforms have qualified their relationship with workers as a coordinated and continuous collaborations (it is the case of the Foodora delivery platform), and yet the platform is still able to pay fees which are far under minimum wage, letting workers earn something like three euros per hour. Therefore, even when it is possible to deem as quasi-subordinate the workers who continuously work on the same platforms, the qualification in terms of quasi-subordinate workers is not per se sufficient to guarantee further protection to those digital workers who would not be able to reach reclassification as employees. The classification as quasi-subordinate workers, conclusively, should not indeed be considered a sort of panacea, and would probably end up creating even more uncertainty.
2.3. Self-employed workers in many cases, unfortunately
The considerations set forth on the qualification of digital workers’ legal status under Italian law suggest that even though the language used in the terms and conditions set by platforms may be seen as “twisted language … [that] merits, we think, a degree of scepticism”, the claim that platforms are not parties of any employment relationships is not that easy to undermine. Even outside Italian law and its narrow notion of employee, the qualification in terms of self-employment laid down in the platforms’ terms and conditions appears convincing de iure condito, although the often-unsustainable consequences of such qualification (with reference to working conditions and occupational stability) may suggest the opportunity to consider the platforms as employers or at least joint employers.
The conclusion that the examined digital workers are in most cases self-employed contractors does not derive from an overvaluation of the contractual label (nomen iuris), which is substantially irrelevant. Such a qualification, instead, is strongly suggested by the circumstance that – safe for pathological cases – they are not actually bound to the directive power of any employer as long as they truly retain the freedom to choose when and where to work.
Such conclusion is coherent also with European law, which does not impose any wider qualification criterion (and probably could not do so, as there is no EU competence on the qualification of employment relationships). In fact, the European Court of Justice individuated the essence of subordination in the circumstance that the worker “acts under the direction of his employer as regards, in particular, his freedom to choose the time, the place and the object of his work”.
The “freedom” of the worker and the presence of a plurality of users suggest that the legal framework of digital work could actually be the triangular scheme – proposed by platforms themselves – of “self-employed work intermediation”, involving three contracts: one self-employment contract between the worker and the user, and two frame intermediation contracts stipulated by the platform with both the worker and the user.
At the beginning of the paper it has been noted how such scheme may deprive “digital” (self-employed) workers of any protection. The following paragraphs will try to analyse more in depth the relationships involved in the triangular scheme proposed by the platforms themselves, in order to verify whether – when reclassification would not be accorded by a judge – the rules governing the specific relationships, as well as those applicable by virtue of the contractual integration between them, may prevent digital workers to fall within an “empty space of law”.
2.4. Mere intermediary or party to the self-employment contract?
If there is a self-employment relationship, we should first ask ourselves who the counterpart of the worker in such relationship is. If we accept the reconstruction operated by the platform, we should say that it is only the time-per-time user, and that therefore not even he/she who works eight hours per day via the same platform can be considered someone “continuously serving a same main client”.
While we have seen that the qualification in terms of self-employment laid down by the platforms may appear convincing (safe for pathological cases), the claim that they just intermediate the provision of transportation services by the users seems more artificial, almost absurd. About Ubers’ activity, labour judges have noted that “Uber does not simply sell software; it sells rides. Uber is no more a “technological company” than Yellow Cab is a ‘technology company’ because it uses CB radios to dispatch taxi cabs” and that it is “unreal to deny that Uber is in business as a supplier of transportation services. Simple common sense argues to the contrary”. Also competition law judges held the platform responsible for carrying a transportation service, and a Barcelona judge requested for an ECJ’s preliminary ruling on the question of the nature of the activity carried out by Uber.
The same functional approach proposed to individuate in the platforms the employer or one of the employers of the digital worker could be useful also in the different perspective of the individuation of the counterpart of the self-employement relationship.
The platform acts indeed as a “cumbersome middleman” and the intermediation contracts are both deeply connected with the self-employment contract. In fact, the intermediation contract sets the frame within which several self-employment contracts are stipulated by the worker and a plurality of clients. Platforms do not only intermediate service; they provide services to users by connecting them to the workers who would actually perform the required activity following the indications set by the platforms themselves.
A good example of the effects of the contractual integration between the self-employment relationship and the frame intermediation contract comes from Chinese case law regarding the platform operator Didi Chuxing, which acquired Uber’s Chinese operations in August 2016. Although Chinese courts were reluctant to accord a reclassification of the “cooperation agreements” signed by the workers, they have in fact imposed some degree of responsibility on the platforms in cases concerning liability for traffic accidents, brought forward by third parties seeking compensation for traffic incidents caused by drivers.
Therefore, as platforms do intermediate, they will respond for the obligations deriving from the intermediation contracts they subscribe with the users and the workers. The mediation contract is also a contractual type regulated by dispositive provisions of the Italian civil code (art. 1754 and ff.). The user-platform relationship shall meet the requirements set forth by consumer law when the user is a physical person, acting for non-entrepreneurial purposes, and even with regard to the worker-platform intermediation contract it would not be out of place to think about the application of consumers protection against vexatious clauses, emphasising the circumstance that the worker acts as a “prosumer”.
However, in the perspective of the aforementioned contractual integration, as platforms set also the rules governing the self-employment relationship, they do also become a party to that relationship or, at the very least, they should still be held responsible for those breaches of the self-employment contract to which they participated, even when the input comes from the user.
We may therefore find in the worker-platform relationship the character of continuity that is missing with regard to the relationship between the worker and the user (entailing the application of quasi-subordinate discipline). In addition, even the most occasional worker may enforce against the platform the rights deriving from the self-employment relationship.
But what rights are we talking about?
3. Self-employed work rights and their sources
In Italy, as in most civil law countries, the discipline of self-employed work contracts is quite gaunt (only seven articles in the Civil Code, art. 2222 to 2228) and it construes the notion of the self-employed worker in negative, by stressing the lack of subordination. Labour lawyers, with some exceptions, have not often focused on such discipline, as they have limited the analysis to the problem of qualification.
The challenge of “digital work”, however, could be an opportunity to develop a new perspective on the protection of self-employed personal work. It does not seem a coincidence that the recent Italian Bill 2233 (infra, § 3.2) contains in the same text provisions on “autonomous non-entrepreneurial work” (first part) and employment provisions “promoting flexibility with reference to the working time and place” (second part). The structure of the bill itself thus confirms that the digitalisation of labour – of standards types and of new forms of labour – requires action on different frontlines.
Moreover, European contract law has indeed developed in the last decades a human dimension in regulating contracts characterised by the imbalance of the parties, on the ground of the interpretative evaluation of the general clause of good faith as well as because of the legislative intervention in the field of consumer law and B2b contracts. In this perspective, it has been underlined that contract law represents nowadays – perhaps even more than labour law itself – “a fruitful field for the ethical evaluation of entrepreneurial behaviours”.
With respect to that apparent “empty space of law” where self-employed digital workers seem to fall, it is important to stress out that some rules would still apply. Contract law general principles (such as the principle of good faith and correctness), would find full application, prohibiting those behaviours that result in the abusive exercise of the rights descending from the contract.
On another level, it would be appropriate to consider the application of the discipline of B2b contracts, and in particular of art. 9, Law 192/1998, which prohibits the “abuse of economic dependency”  of a small firm towards a main client. Although such provision is contained in a Law regulating the “sub-supply” commercial relationships, the Italian Supreme Court clarified that article 9 has a wider scope than the other provisions of that Law, as it constitutes a “general clause”, applicable to any contractual relationship in which an abuse of economic dependence may occur. On this ground, it seems reasonable to extend the application of the prohibition of the abuse of economic dependency to self-employment relationships, in order to protect, at the very least, autonomous workers from suffering unilaterally and arbitrary decisions made by their counterpart.
The combination of the few Civil Code provisions regarding self-employment, together with the general principles of contract law and the extensive interpretation of the B2b contracts rules, may result in a discipline capable to provide some protection against some of the critical issues raised by platform-mediated work, as the following paragraphs will try to argue.
3.1. Some consequences under private law
Most platforms retain the power to exclude the worker from the use of the platform, by deactivating his/her account. If we consider digital workers as employees such deactivation may be deemed as a dismissal, and would therefore need to comply with national and European provisions requiring the dismissal to be justified. Self-employed workers, instead, do not enjoy the same warranties. The relevance of the problem of “deactivation power” emerges if only we take into account the circumstance that one of the conditions contained in the Cotter v. Lyft settlement proposal provided for the enforcement of a grievance process heard by an arbitrator to be undertaken before account deactivation.
Deactivation, actually, does look more like the termination of the intermediation contract than like the termination of the self-employment relationship that is framed within it. In this perspective, the specific provision about the termination of the self-employment contract provided for by art. 2227 of the Civil Code, providing for the right to terminate at will the relationship by paying the worker a compensation, does not seem any useful. If we consider deactivation as the termination of the intermediation frame, on the contrary, we could usefully recur to contract law general principles and B2b contracts regulations to syndicate its legitimacy.
Two decisions from Southern Italy regarding the famous online auction and shopping website Ebay may represent a good example. Both decisions, rendered in the contest of the special “urgency” proceeding provided for by art. 700 of the Civil Procedure Code, ordered the Company to re-activate the accounts of two sellers who had been de-activated due to negative feedbacks.
The first decision deemed unlawful the deactivation under the general rules on contract termination set forth by the Civil Code (art. 1454 and ff.), considering the mere presence of low feedback ratings as not sufficient to demonstrate the existence of a serious breach of the contract. The judge considered that a clause allowing resolution only due to negative feedbacks would be a vexatious clause requiring double subscription (missing in the case) for its enforceability. However, it has been noted that even in the presence of a second subscription, the clause would still be void under art. 9 L. 192/1998, as it realises “an abusive imposition of unjustifiably vexatious conditions”.
The second decision seems more aware of the social and economic dimension of the problem and gave relevance to the oligopolistic structure of Ebay’s on-line marketplace. The judge recognised the existence of the so-called periculum in mora (necessary to access to the urgency proceeding) because “the exclusion from Ebay does not only produce some lost clients, but excludes a micro-business from the market itself”. However, the judgement explicitly excluded the application of consumer protection law and of Law 1992/1998, on the ground that there was no “introduction of the micro-business in the productive process of a main client”.
What is interesting about the two aforementioned decisions is that, even moving within the field of general contract law, they are still able to grant the weak party of the relationship a real protection, a sort of reinstatement, reaching an effect that recalls the traditional sanction against unjustified dismissal. It is a revolutionary conclusion, as it goes even further the – yet quite revolutionary – conclusions reached in the Renault Case, where the ascertainment of an abusive termination of the contract entailed only compensatory remedies.
Many of the considerations resulting from by the aforementioned decisions deal with the reputational systems adopted by some platforms. With regard to digital workers, it would be easier to invoke the invalidity under art. 9 Law 192/1998 of vexatious clauses granting termination at will powers to the counterpart, as there is nothing ancillary to the platforms business in their activity. In this perspective, it is possible to give a partial answer to the question is account deactivation a new form of dismissal?Without the reclassification of the relationship, there is no room for the application of statutory protection against dismissal. However, it is still possible to prevent the unjustified exclusions of workers from the platforms. Low ratings can bring to deactivation only when they derive from a seriously neglect conduct of the worker and a minimal procedure to allow the worker to defend himself/herself shall be accorded in any case under the general clause of good faith.
The issue is strictly connected with the problem of the control that the platform is capable to exercise on the execution of the performance, even by delegating it to users by means of reputational systems. Such control, in fact, may be compatible with the self-employed nature of the relationship, but only as long as it remains a control on the result of the work – in order to guarantee minimum standards of quality and safety – and not on the execution of the worker’s performance. Nonetheless, the border between these two objects of the platforms’ control can indeed be quite evanescent in many cases.
Certainly, however, personal ratings should not depend on the amount of time the worker devotes to the tasks delivered via app, coherently with the alleged self-employment status. Should the reputational system “punish” dormant workers, they would be able to react invoking an employee status, as they would end up to be at the disposal of an employer (as Judge Snelson noticed, quoting Milton, “they also serve who only stand and wait”). Nevertheless, at the same time, they would also have the possibility to invoke their self-employment status in order to prevent the platform from affecting their rate or to “dismiss” them without a concrete reasonable cause. Although it is clear that the acknowledgement of an employment status would bring much more benefits, it may also be useful to provide the worker with a “second bullet”, in a context characterised by uncertainty and by the malleability of employment tests.
Although the case of the termination of the contract through account deactivation seems paradigmatic, a similar approach could be adopted to ascertain the legitimacy of the clauses allowing the user to refuse the acceptance of a performed task, without providing payment to the worker, as well as to question the legitimacy of the exclusivity clause that some platforms insert in their general conditions.
In the first case special provisions on self-employed work regarding the right to receive due compensation may apply. Art. 2227 of the civil code provides the client with the right to terminate the self-employment relationship when the task has been partially executed, “compensating the worker for the expenses, for the performed work and for his loss”. Case law stated that the “loss” under art. 2227 c.c. is constituted by the full price of the agreed performance. The special provision on self-employed “intellectual performances” (art. 2237 c.c.) provides the worker with the right to be compensated for the expenses and to be paid for the performed work, which will be quantified “with regard to the utility deriving to the client”. Under both these regulations, which prohibit termination without compensation when the performance is not yet completed, it seems that the clause allowing the client to refuse a completed task should be, a fortiori, not enforceable under Italian law.
Even the exclusivity clauses may be deemed unlawful under Italian law. In the first place, as they result in “restrictions to the freedom to contract with third parties” they certainly require double subscription under art. 1341 c.c.. In the second place, it could be appropriate to consider such clause as vexatious under B2b statutes, and thus void under art. 9 L. 128/1998, adopting the same reasoning proposed with regard to the remedies against account deactivation.
It seems therefore possible to address some crucial issues regarding the protection of digital workers also through the application of non-employment regulations: the general rules of contract law, the specific rules on self-employment and the regulations regarding B2b contractual relationships. However, the recourse to civil law principles and regulations is still far from being a satisfactory solution, as it presents all the weaknesses of an interpretative-only solution and leaves many critical points unsolved.
3.2. Towards the development of a Statute for self-employed (digital) workers
An answer to the absence of an exhaustive discipline on pure self-employed work comes from the recent Italian Bill AS 2233/AC 4135 (definitively approved by the Senate on the 10th of May 2017) containing “protective provisions on self-employed non-entrepreneurial work”.
As it has been underlined, the legislator tried for the first time to construe a discipline of self-employed work based on the acknowledgement of its social and ethical value, rather than on the prejudice that it hides actual employment relationships.
The Bill introduces several warranties for self-employed workers, ranging from the protection against payment delays (art. 2), to tax benefits (art. 8 and 9), to the access to formation and collocation services (art. 10) and public procurement (art. 12), to social security benefits such as (unpaid) maternity leave, sick leave and injury leave (art. 7, 13 and 14).
In the perspective of digital self-employed work, one of the most important provisions is set up by art. 3 (4), which explicitly provides for the application to self-employment relationships of the aforementioned art. 9 L. 192/1998, thus removing the uncertainties of interpretative extension. In addition, art. 3 (1) specifies that the clauses “granting the client the power to unilaterally modify terms and conditions and, where the self-employment relationship is characterised by continuity, to terminate the relationship without notice” are vexatious and thus void.
It is not clear whether the invalidity of the vexatious clause would lead to real remedies (such as the re-constitution of the relationship or the disapplication of the clause), as the Bill provides that the worker would be entitled to receive reparation for damage and loss (art. 3 (3)). However, it seems possible to interpret such provision as granting reparation for the damage and loss related to the enforcement of the vexatious clause, without excluding the possibility of the restoration of the status quo ante at the request of the worker.
A more specific attempt to regulate the provision of personal services in the gig-economy is represented by the Bill 3564, containing “provisions on digital platforms for the sharing of goods and services and provisions promoting sharing economy” (so-called Sharing Economy Act). Although the Bill reflects somehow the misunderstanding that the gig economy represents an aspect of sharing economy, and its purpose is mainly to promote sharing economy – with an eye to the tax increase that may derive from its development – the Bill contains some provisions which may be extremely relevant for the purpose of granting a fair treatment to platform-mediated workers.
Art. 4 of the Bill provides that the platform owners shall adopt a written policy, subject to the Competition Authority’s approval, including the contractual terms and conditions between the platform and its users. The Bill provides for a list of clauses penalising the “user-operator” (broad label that seems to include also those that we have called “digital workers”), which are expressly sanctioned with invalidity. In particular, platforms terms and conditions cannot “a) burden the user-operator with any kind of exclusive obligation; b) allow the control on the execution of his performance, not even through hardware or software systems; c) determine compulsory fees for all users; d) allow the exclusion of the user-operator from the platform or penalise him in the presentation of his offer without serious reasons; […]h) forbid the user operator from criticising the owner of the platform” (art. 4 (2)).
Although the Bill is meant to regulate all kinds of sharing economy activities, without a specific labour law focus, the provisions that we have just examined – promoting transparency and fairness in the platform’s management – should be welcomed as they represent a consistent step forward in filling up that “empty space of law” where digital workers seemed to fall.
The frustrations raising from the difficulty in applying statutory employment law to digital workers should not lead to the misunderstanding that no protection can be found outside that domain. The “lightness” of intermediated self-employed work remains “unbearable”, but some attempts to make it heavier may be crowned with success.
Through the valorisation of contract law principles and regulations, we may already be able, at an interpretative level, to address some of the issues raised by platform-mediated self-employed work. The legislative perspective of implementing new sets of rules for “pure” self-employed workers and digital “users-operators” may also give further answers to the exigencies of digital workers.
Some crucial points, however, remain unresolved. In the first place the problem of low wages, which represents a constant of platform-mediated work, with workers being paid much less than minimum wages set by applicable legislation or collective agreements. Two doctrinal proposals in this field deserve to be highlighted, as they move from the awareness of the difficulties in reaching reclassification for “digital workers”, coherently with the structure that has been adopted in this paper.
The first proposal is to extend the scope of minimum wage provisions provided for by law and collective agreement also to self-employed workers, by means of a specific intervention (also in the field of European competition law).
The second proposal, on the other hand, suggests an interpretative extension to (self-employed) platform workers of the principles provided for by Directive 2008/104/EC on Agency Work, and in particular of the equality principle enshrined in art. 5 (1) of the Directive.
Rebus sic stantibus, however, it seems extremely difficult to address the issue of low wagesby recurring to the existing instruments of law. The constitutional principle of a “proportionate and adequate salary” (art. 36 Italian Constitution) has been repeatedly declared inapplicable to self-employed workers, and the only provision granting an “equivalent” wage to quasi-subordinate self-employed workers has been abrogated by the Jobs Act reform.
Many other issues, moreover, such as the risks of self-exploitation and exploitation of child labour and the difficulties in pursuing effective collective representation for an atomised working force cannot find as of today a satisfactory solution outside the field of statutory employment law.
The challenge to avoid “digital work” from bringing to a “nosedive” labour law standards, therefore, calls for a political reflection to be conducted both at national and at supranational level.
* The paper is the revised and integrated version of the draft paper presented at the 15th International Conference in Commemoration of Professor Marco Biagi, Digital and Smart Work, Modena, Marco Biagi Foundation, 20-21 March 2017. A reduced version of the paper will be published in the conference proceedings. References made in the paper to Italian courts’ decisions and scientific reviews follow the editing criteria adopted by the Rivista italiana di diritto del lavoro.
H. Simões Gomes, Justiça do Trabalho de MG decide que motorista tem vínculo empregatício com Uber, in globo.com, 14 February 2017; J. Weinberg, Gig News: Brazilian Judge Finds Uber Driver is Employee, in Onlabor.org, 14 February 2017.
Reuters, Justiça do Trabalho de MG decide que motorista da Uber não é funcionário da empresa, in globo.com, 24 May 2017.
V. De Stefano, The Rise of the “Just-in-Time Workforce”: On-Demand Work, Crowdwork, and Labor Protection in the “Gig Economy”, in Comparative Labor Law & Policy Journal, 2016, vol. 37, n. 3, p. 474; E. Dagnino, Il lavoro nella on-demand economy: esigenze di tutela e prospettive regolatorie, in Labour Law Issues, 2015, 2, p. 90, who observes that the main difference is that the former involves an “on-demand virtual workforce” while the latter involves an “on-demand mobile workforce”. In the perspective of a unitary approach also J. Prassl & M. Risak, Uber, Taskrabbit, and Co.: Platforms as Employers? Rethinking the Legal Analysis of Crowdwork, in Comparative Labor Law & Policy Journal, vol. 37, n. 3, p. 619, who however use the term “crowdwork” for both the aforementioned types of work-on-demand.
M. Weiss, Digitalizzazione: sfide e prospettive per il diritto del lavoro, in DRI, 2016, n. 3, p. 662.
A. Aloisi, Commoditized Workers: Case Study Research on Labor Law Issues Arising from a Set of “On-Demand/Gig Economy” Platforms, in Comparative Labor Law & Policy Journal, 2016, vol. 37, n. 3, p. 688, where the A. indicates four key variables that may differ from platform to platform (means of exchange, system of payment, population of the users and workers’ status).
Aslam, Farrar et al. v. Uber B.V. et al. Among the first commentators: G. Davidov, The Status of Uber Drivers: A Purposive Approach, forthcoming in Spanish Labour Law and Employment Relations Journal; I. Lloyd, Uber Drivers in London: “To Be Or Not To Be” An Employee?, in Computer Law Review International, 2016, n. 6, p. 161.
Ivi, p. 29, n. 8.
 B. Bergvall‐Kåreborn & D. Howcroft, Amazon Mechanical Turk and the Commodification of Labour, in New Technology, Work and Employment, 2014, vol. 29, n. 3, p. 213; Aloisi, Commoditized Workers, cit. p. 653.
Aslam, Farrar et al. v. Uber B.V. et al., n. 97. Also Advocate General Maciej Szpunar, delivering his Opinion in the Uber Case pending before the ECJ (infra, n. 75)noticed: “the company may very well provide its services through independent traders who act on its behalf as subcontractors” (§ 54).
 Among the most recent, Cass. 8 November 2016, n. 22658; 3 October 2016, n. 19701; 19 September 2016, n. 18320, all in De Jure.
P. Tullini, C’è lavoro sul web?, in Labour Law Issues, 2015, vol. 1, n. 1, p. 9.
A. Perulli, Economically dependent / quasi-subordinate (parasubordinate) employment: legal, social and economic aspects, European Commission 2003, p. 6.
 With reference to crowdwork, A. Felstiner, Working the Crowd. Employment and Labor Law in the Crowdsourcing Industry, in Berkeley Journal of Employment and Labour Law, 2011, 32:1, p. 168, who pointed out as platforms deliberately decided to take advantage of the fact that “there were virtually no cases, and few indications in the legal literature as to how courts might approach regulation of the ‘cyberspace workplace’”.
Tullini, C’è lavoro sul web?, cit., p. 8.
Cotter et al. vs. Lyft Inc., Case No. 13-cv-04065-VC, Order denying cross-motions for summary judgement (California Northern District Court 11 March 2015).
Ivi, p. 13. The judge consequently denied issuing a summary judgement, referring the case to a jury.
M. Cherry, Beyond Misclassification: The Digital Transformation of Work, in Comparative Labor Law & Policy Journal, 2016, vol. 37, n. 3, p. 577, providing an in-depth analysis of U.S. litigation.
S. Kessler, The Gig Economy Won’t Last Because It’s Being Sued To Death, in fastcompany.com, 17 February 2015; C. DeAmicis, Homejoy Shuts Down after Battling Worker Classification Lawsuits, in recode.net, 17 July 2015.
 As in the aforementioned Cotter v. Lyft case, settled on 27 January 2016 for 12 million dollars, and in O’Connor et al. v. Uber Technologies Inc. et al., settled on 21 April 2016 for almost 100 million dollars. Cherry, Beyond misclassification, cit., notes that “the result is ultimately disappointing for those who saw this as a case that would most likely set a precedent”.