“Can You Complete Your Delivery?” Comparing Canadian and European Union Legal Statuses of Platform Workers

In December 2021, the European Commission proposed a directive creating five criteria for the presumed classification of platform economy workers as salaried employees. The issue is timely, of course, as the digital organisation of work contin‐ ues to grow rapidly. Our article contrasts the merits and limitations of this initiative to the Canadian experience concerning so‐called independent contractors in the platform economy. In fact, Canadian labour law has long recognised a third status of workers—dependent contractors. It permits collective bargaining, while platform workers remain autonomous, notably for tax purposes. Immediately, the striking similarities between the European Union’s five criteria and judicial tests applied by Canadian labour tribunals seem to indicate that both entities are moving in the same direction. However, the federal structure of labour law in Canada and the single market’s social dimension also pose important challenges regarding the uniform implementation of new protections. Based on recent fieldwork in Toronto, and as the European Union directive moves into the approval and implementation stages, our article addresses the research question of how basic labour rights in the platform economy progress similarly (or differently), and which actors are driving the change on each side of the Atlantic. We argue that this policy field provides labour market actors with opportunities for “institutional experimenta‐ tion” navigating the openings and limitations of federalism.


Introduction
In December 2021, the European Commission proposed a directive creating a set of five criteria intended to provide a unified basis for the presumed classification of platform economy workers as salaried employees within the single market (European Commission, 2021a). The issue is crucial, as app-based organisation of work continues to grow rapidly in the retail, food delivery, and transportation sectors. This article compares the ongoing progress and substantive developments of the EU directive to the recent Canadian experience concerning so-called independent contractors in the platform economy.
In fact, Canadian labour law (both federal and provincial) has long recognised a third status of workers-dependent contractors. It permits collective bargaining and some health and safety coverage, while platform workers remain autonomous, notably for tax purposes. The striking similarities between the European Commission's five criteria, as presented in the draft directive, and judicial tests commonly applied by Canadian labour tribunals seem to indicate that both entities are moving in the same direction. They are both attempting to avoid workers' misclassification and regulate employment within the platform economy.
The federal structure of labour law in Canada and the importance of subsidiarity in the EU pose important challenges when applying new protections uniformly and implementing said policy initiatives. Canadian labour law is indeed a patchwork of one federal and 10 provincial legislations. Similarly, social protections for workers do not fall within the exclusive jurisdiction of the EU. Even directives adopted under the mantle of the four freedoms need to proceed along the winding road of member-state implementation, risking divergence from the norm.
This article's research question explores how basic labour rights in the platform economy progress similarly (or differently), and which actors are driving the change on each side of the Atlantic. In doing so, the article contributes to long-standing debates about policy innovation in multi-level-governance systems. In particular, we are interested in the role of spillover (Haas, 1958;Niemann, 2021) whereby higher-level governments are pushed towards legislation by labour market actors (e.g., large multinationals, employer organisations, and trade unions) or policymakers on lower levels. By making contradictory policy demands and proposals, legislative action at the highest level may be required in order to maintain policy coherence. This then poses the question of "subsidiarity" (Endo, 1994) in federalist systems, unearthing potential sources of resistance when implementing directives from the highest level. A different body of literature anchored in sociological institutionalism insists on the role of isomorphism (DiMaggio & Powell, 1983) within intertwined jurisdictions. Different levels of government thereby "mimic" (Sisson, 2007) policy initiatives that are seen as novel or successful, which may encourage (or hinder) policy experimentation. An additional focus will be on the role of labour market actors versus legislators and courts when approaching this multi-level game of policymaking.
The multidisciplinary approach, combining labour law, policy research, and labour market sociology, is meant to enrich the discussion. It is not meant to create uniformity or an overarching framework where none exists. If anything, the adaptation to a relatively new phenomenon in the labour market, such as the spread of platform work and app-based services, should be met with a flexible theoretical framework. One that comes to mind is "institutional experimentation" (Ferreras et al., 2020), which is defined as a process by which labour market actors assume specific roles in shaping institutional change within larger policy frameworks, resources and contingencies. This article argues that neither actorbased innovations alone nor a purely legal approach can result in effective protections for platform workers. In our case, the heterogeneity of labour law frameworks within two multi-level governance structures, Canada and the EU, provides such an open field for institutional experimentation.
Answering why the EU and Canada provide for a fruitful comparison of platform workers' legal statuses hinges precisely on the interplay of labour market actors and institutions. The multi-level governance structure of both polities, albeit constitutionally quite different, opens up space for actor-based innovations where more homogeneous nation-states tend to compress policy innovation into established path dependencies (Pierson, 2000). Institutional "layering" (Streeck & Thelen, 2005), "bricolage" (Crouch, 2007) by social actors facing blocked or alternate paths, and the "ambiguity and agency" (Mahoney & Thelen, 2010) such a setting creates have long been concepts used to explain actor-induced institutional change over time. The cases of platform workers in Canada and the EU provide two very promising cases for such an analysis.
By proceeding in this way, the article contributes first and foremost to a better understanding of labour law and labour market actors within comparative political science, focusing on the role that creative experimentation and policy advocacy plays in shaping institutions. The article also speaks directly to practitioners on each side of the Atlantic. By highlighting the opportunities and challenges of organising platform workers in the delivery and transport sectors, and by contextualising them in the complexities of federal labour law, practitioners in both geographical locales (and beyond the subsectors covered by our fieldwork) can draw important lessons for improving the plight of millions of platform workers, be it in Canada or the EU.
The article progresses as follows: After a brief overview of definitions and the empirical basis for the article (Section 2), we will present the fundamentals of multi-level employment law, as it is applied to platform workers in the two cases (Section 3). Then, we will describe our findings from the Canadian fieldwork (Section 4) and discuss legislative initiatives in that country (Section 5). Crossing the Atlantic, we will then elucidate national initiatives on regulating platform work in Europe (Section 6), before turning our attention to the EU directive itself (Section 7). In Section 8, we will discuss the drivers of similarities and differences for the variable progress shown in the two cases, before ending with some conclusions for theory and practice (Section 9).

Empirical Basis of the Article and Key Definitions
Embarking on an analysis of institutional change and the role of social actors in a setting of multilevel governance necessarily comes with its own pitfalls. Firstly, definitions of key institutional concepts may not concur. We have used the term "multi-level governance" (Scharpf, 1999) to describe labour law in both Canada and the EU. We, of course, realise that the European Union is more of a "supranational polity" (Hix, 2007) or a confederation, based on (some) upwards delegation of jurisdiction while relying on national-level implementation and respecting the principle of subsidiarity. In comparison, Canada is a constitutional federation with relatively clearly divided responsibilities in the field of labour, with each of the institutional levels (federal and provincial) overseeing implementation independently (see Article 92 in the 1876 Constitution Act; Minister of Justice, 2021).
Secondly, we have applied the legal constructs of "salaried employee" and "independent contractor" uniformly while realising that the notions imply somewhat different criteria on each side of the Atlantic. The resulting in-between statuses are even trickier. For the purposes of this study, we distinguish two. Under "dual-status" workers, we include those who find themselves recognised vis-à-vis a designated employer for most of the purposes of labour legislation, while retaining the status of "independent contractor" notably for purposes of taxation and contract law. With "third status," we mean a detailed hybrid, by which only some, limited parts of labour law apply to them. The variations of such an in-between model are too plentiful to enumerate them all. Some allow for collective bargaining of wages and working conditions in a separate negotiation framework but do not include social security protections. Others focus on occupational health and safety (OHS) and workers' compensation schemes, but without extending the right to collective representation.
Finally, given the multiple forms of app-based work and the various incarnations of the gig economy, we must focus on the precise sector that our study covers by analysing applicable employment law, labour policy, and actor-based innovations in the context of in-person services mediated by platforms (digital apps acting as intermediaries between customers, service providers, and workers). More specifically, we are only examining the food delivery and transportation subsectors. They are of particular interest, as they have seen tremendous growth in numbers, both absolute and relative to the more traditional service sectors against whom they now compete. While we appreciate that this limits our findings-to two very dynamic subsectors with relatively low-skill workers-and generalisations with other subsectors of the gig economy might be difficult, we believe that experimentation can best be studied in a context of disruption and rapid growth.
Based on recent fieldwork concerning the organisation of food-delivery and transportation platform workers in Toronto, as well as some expert interviews on each side of the Atlantic, we follow a largely inductive epistemology, drawing inferences and developing implications for social and political theories as we progress. In Canada, a total of six semi-structured interviews each of approximately 45 minutes were held at different levels (local organisers and national union representatives) and with interviewees from different occupational backgrounds (e.g., riders and drivers). In Europe, we conducted five expert interviews covering three different countries as well as the EU as a whole. They varied from 30 minutes to an hour in length (for the complete interview list, see the Supplementary File).
To circumscribe the legal lay of the land on each side of the Atlantic, we completed an analysis of applicable legislation and policy documents. In Europe, we analysed seven national legal frameworks (Belgium, Denmark, France, Germany, the Netherlands, Norway, and Spain) and three supranational texts (the European Commission, Council, and Parliament). In Canada, we concentrated on federal, Ontario, and Quebec legislation. We also executed extensive documentary research, analysing over 30 texts stemming from seven different, independent media sources (two Canadian and five European), including a six-part podcast series by the Toronto Star. We also obtained trade union documentation (nine Canadian and six European policy documents as well as 11 Canadian and five European press releases found on the respective organisations' websites). Additionally, we also analysed three Canadian press releases issued by the multinationals Uber and Foodora.
Before exploring these extensive information sources in more detail, the next section will embark on an overview of the institutional and legal framework covering platform workers in the EU and Canada. The section is arranged by government level and labour policy field to explore labour standards such as health and safety, working time and minimum wage, collective bargaining, and trade union accreditation.

Platform Workers in the Context of Multi-Level Employment Law
At least 28 million workers in the EU are currently working for digital labour platforms, a number forecast to rise to 43 million by 2025 (European Commission, 2021b). A Canadian study (Action Canada, 2021) showed that 28% of Canadians draw some form of income from digital platforms, with it being the main income source for one in four. This would put the total number of Canadian platform workers at roughly 2.6 million at the time of the study, likely to be even higher today.
Platform workers, especially in the transportation and food delivery sectors, do not typically choose their status as autonomous workers deliberately. As our interviews revealed, above all, they need "a job" without too many entrance requirements. The rapidly expanding platforms provide them. The abundance of young, migrant, and often racialised workers in Canada are readily absorbed into this segment of the service industry. This seems to apply similarly in Europe (Altenried, 2021). The lack of protections related to not being a salaried employee-albeit intimately part of the platforms' business model-often comes as an afterthought to workers.
The stark dichotomy between salaried employees and independent contractors is put to multiple tests in this new world of highly mobile labour. Platform work, while technically considered to be independent, often falls between the cracks. For the worker, many of the benefits of being truly independent are absent and protections (linked to a stable salaried employment status) are patchy. In practice, social actors and policymakers have thus toyed with various forms of a "dual status." This would entail giving platform workers certain rights from both categories, as salaried employees of clearly identifiable employers (e.g., for collective bargaining purposes) and as independent contractors (e.g., for tax purposes). This contrasts with classifying them into various forms of "third status." Such a classification would take away the independence of being a contractor and provide only a limited set of the rights habitually granted to salaried employees, thus effectively they become neither independent contractors nor salaried employees.
Complicating this conundrum is the decentralised nature of labour law in both of the jurisdictions being compared. In Canada (see Figure 1), over 90% of all employment is considered provincial jurisdiction under the Canadian Constitution (Statistics Canada, 2021). Federal labour legislation (Minister of Justice, 1985) primarily covers banks, interprovincial/international transportation and shipping, telecommunications, and the federal civil service (including state enterprises). Only the category of "postal and courier services" might provide for some application of federal labour legislation to in-person platform work. Thus, most platform work is regulated by provincial labour codes.
The federal labour code, as well as those of nine of the provinces (all influenced by jurisprudence stemming from the Anglo-Canadian common law tradition), already recognises a third status as "dependent contractors" (Minister of Justice, 1985, Art. 3.1c). While the province of Quebec, representing roughly 20% of the Canadian workforce, does not. Its civil law tradition opted to create distinct legal frameworks for various forms of "dual statuses" (e.g., for artists and childcare operators) instead.
Legislation on workplace accidents and occupational diseases in the provinces is very similar (the federal legislator has not created its own workers' compensation scheme). Therefore, federal employees follow their respective provincial legislation-a no-fault collective insurance paid for by employers. Henceforth, and unless the workers contribute themselves to said schemes on a voluntary basis, this generally excludes independent and dependent contractors. The apparent (mis)classification of platform workers adopted by each province, thus becomes a critical weakness for workers' effective protection against workplace accidents.
In the EU (see Figure 2), only broad minimum standards can be adopted under the mantle of the acquis communautaire aiming at protecting fair competition. Such has been the case, under Art. 153.1b, in the fields of working time and paid annual leave. While social protections, such as employment insurance and rules around termination of employment (Art. 153.1d), require unanimity among member states and are thus next to impossible to harmonise. Right of association and wage setting (Art. 153.5) continue to fall entirely under national jurisdiction (Consolidated version of the Treaty of the European Union, 2012).
OHS regulation exemplifies the most integrated policy field-a framework directive, specific directives about issues such as protective equipment and contaminants, and a multitude of binding standards. It also has its own enforcement agency (European Agency for Health and Safety at Work, 2023). Together, these elements have created an evenly and directly applicable regulation, thereby providing a common floor for workers across the EU. The aim here is to prevent the cutting of corners on health and safety at work from becoming grounds for competitive advantages or stark differences in OHS approaches thus limiting the four freedoms (free movement of goods, services, capital, and labour) within the single market in some way. The problem, of course, is that independent contractors are not always covered; thus, a policy initiative for misclassified platform workers is crucial.
After having laid out the main legal dilemmas created by the classification issue, the next section will bring our discussion into the details of the consequences of misclassification in the day-to-day reality of platform-based work in Canada. It will also discuss the

Risks and Benefits for Platform Workers as Revealed by Our Canadian Fieldwork
On substance, what are they complaining about? When determining misclassifications, both the proposed EU directive and the Canadian jurisprudence insist on subordination and control over working conditions as key elements. As revealed by workers' testimony before the Ontario Labour Relations Board (OLRB), in the case of Foodora, food-delivery workers are subject to intense surveillance-ranging from tight timelines for pick-ups and deliveries to algorithm-imposed disciplinary measures. They also have no possibility to negotiate their remuneration; rather, they have predetermined fee schedules by distance and/or type of service. This creates an "economic dependency" according to the tribunal's ruling (Canadian Union of Postal Workers v. Foodora Inc., 2020). While some elements of the Canadian five-step judicial test, which will be presented later in more detail (Table 1), may point to a status as independent contractors (e.g., the possibility to work for different food delivery platforms at the same time and the risk of economic losses stemming from long wait times at a restaurant), the OLRB ultimately evaluated the link of subordination and dependency to be more significant. With respect to the obligation to provide one's own equipment, such as a mobile phone with a data plan, a car or bicycle, and protective gear (and, in the case of Foodora, even an obligation to acquire the emblematic, fuchsia-coloured thermal bags), these obligations may also apply to salaried employees under many Canadian labour codes and do not by themselves permit a classification as independent contractors.
As revealed by our interviews with food delivery riders, the most upsetting problem that they faced in the Canadian example is the lack of respect from a relatively anonymous employer. This is exemplified by employers intervening through messenger chats (rather than face-to-face communication) as well as severe health and safety concerns for the workers. Delivering meals on bikes frequently leads to work-related injuries, ranging from road accidents, such as dooring by drivers exiting their parked vehicles and bike tyres becoming trapped in potholes. Especially in the depths of a Canadian winter, both deliveries by bike and by car are replete with safetyrelated issues-repeatedly stressed by our interviewees. A further important safety issue is the absence of protection from harassment and even sexual violence, reported by female delivery workers during our fieldwork.
As previously mentioned, unless they make voluntary contributions for themselves, independent contractors are not covered by Canadian workers' compensation schemes. Nor are their employers required to take preventative measures to ensure a safe work environment for them. A symptomatic situation reported by a Toronto-based rider, who broke his arm from a fall due to a tramway rail, inspired the title of this article. When reporting his accident to the dispatch over the app, the manager's initial reaction was: "Are you still going to be able to complete the delivery?" (Gebert, 2021). This became a rallying cry for the Foodora worker unionisation drive in that city. Severe safety problems, coupled with a flagrant lack of respect for the workers, have propelled the issue of their independent contractor status to the fore catching the attention of traditional labour market actors, such as national Canadian trade unions.
Most Canadian labour codes reserve collective representation and industrial action for workers who are classified as salaried employees or dependent contractors (Gouvernement du Québec, 1964;Government of Ontario, 2000;Minister of Justice, 1985). To access those rights effectively, however, trade unions need to organise them first. This setup provides little to no incentive to do so, once they have been effectively misclassified by their employers as independent contractors.
In the history of collective action and labour organisations, however, the status of autonomously executed work is not necessarily an anathema to collective action and social protection. As a case in point, one of the first trade unions in North America (the shoemakers of Boston) was founded in 1648 upon an association of skilled tradespeople who were seeking to negotiate their fee schedule collectively (Commons, 1909). After a period of prohibition, presumably because it would infringe on antitrust principles (Cox, 1955), many trade unions are still intimately linked to skilled trades. This is especially so in the construction industry, and thus they do have a historical repertoire linking them to individuals identifying as workers but acting as small entrepreneurs for other purposes. This dual identity dubbed "craft unionism" (Perlman, 1922) might very well apply to workers in the gig economy as well.
Canadian legislators have sometimes reacted by enshrining such a dual status into law. For example, workers in the arts, media, and entertainment industry (e.g., actors, screenwriters, dancers, and singers) are considered both independent contractors and workers who are entitled to collective bargaining and basic labour protections (Minister of Justice, 1992). Working on multiple television or filming sets at the same time and signing individual contracts for their services, they are nevertheless organised by major Canadian trade unions and negotiate basic protections and minimum fee schedules with recognised employer associations. Such arrangements are not uncommon, the so-called academic trade unions in Scandinavian countries function in a similar way (Logue, 2019). Also, the Canadian province of Quebec recently allowed home-based childcare providers (while remaining small enterprises for fiscal purposes) to negotiate their fees collectively with the Ministry of Families (Gouvernement du Québec, 2009).
Anglo-Canadian common law jurisprudence and labour codes have instead provided for a "third status," that of "dependent contractor." It entails that autonomous workers must be able to access collective representation and social protection if they can satisfy certain legal tests establishing their subordination to (and dependency on) a presumed employer. This is precisely the scenario of the previously introduced Foodora example. The OLRB, after applying the five-step test regarding subordination, pay schemes, discipline, and control, concluded that the couriers and drivers had been misclassified by their employer and should benefit from unionisation and other forms of collectively negotiated protections (Canadian Union of Postal Workers v. Foodora Inc., 2020). The situation bearing the closest resemblance in Europe, is the "worker status" for employment within the platform economy that predominates in the UK (Rogers, 2019).
In the debate over how to improve the plight of app-based delivery workers, one may ask whether such reclassification is the way forward, or whether a "third status" is perhaps a third rail-zapping a vast variety of protections in exchange for a more limited set of workplace rights. That debate is justified because the salaried employee status remains the most legally binding guarantee for economic and social rights. The problem of inclusion in (or exclusion from) the OHS frameworks of their respective jurisdictions is a case in point: While the OLRB was competent to require union certification, it did not have the mandate to require inclusion into the workers' compensation framework. However, fearing civil lawsuits over work accidents, at least one platform has since elected to contribute to the Ontario workers' compensation scheme on a voluntary basis.
The multiple challenges experienced by platform workers in Canada thus beg two questions: Why have the legislators at the federal and provincial levels not intervened? And what precedent (if any) would apply to their situation? That discussion will be the subject of the next section, which includes a presentation of the role Canadian social actors play in policy innovation.

The Canadian Experience on Policymaking in the Field of Labour Law
Achieving better coverage for Canadian workers through basic labour protections is hindered by the county's multi-level governance structure in labour law-similar to the EU's. Instead of a uniform approach, labour market actors, legislators, and workers have been experimenting with various third and dual statuses for non-standard workers. In other fields of social policy, however, we can see mimicking of successful efforts at the provincial and federal levels. For instance, proactive pay equity legislation first introduced in Quebec in 1997 was finally incorporated at the federal level in 2018. This was due, in large part, to the pressure exerted by prominent public sector unions such as the Canadian Union of Public Employees and the Public Service Alliance of Canada.
Strikebreaker legislation and card-check certification have seen a similar ebb and flow between jurisdictions. For example, Ontario created (1990), then abolished (1995) card-check accreditation. After decades of lobbying by the Canadian Labour Congress, strikebreaker legislation is now being proposed at the federal level, while card-check legislation was only briefly repealed federally between 2014 and 2017. Quebec created and retained both since 1982. However, being subject to the changing politics on both jurisdictional levels, Canada (like the EU) has been unable to provide uniform protections.
In the aforementioned case of the Canadian Union of Postal Workers seeking accreditation for Foodora couriers and drivers in Mississauga and Toronto, the weaknesses of a purely "actor-based" experimentation with existing legislation become evident. Shortly after the OLRB ruled in favour of the Canadian Union of Postal Workers, Foodora exited Canada, thereby leaving the entirety of the Canadian market for app-based food delivery unorganised once again.
A separate but related case involved Uber Eats drivers contesting their dispute settlement scheme with the company. The contested mechanism referred to Canadian disputes in arbitration in the Netherlands. Starting in 2017, the case wound its way through the labour courts. After a final setback before the Supreme Court of Canada, the company agreed to substitute its previous arbitration practice with a voluntary representation scheme with the United Food and Commercial Workers trade union acting as an official interlocutor (Uber Technologies Inc. v. Heller, 2020). The United Food and Commercial Workers is now recognised to represent Uber and Uber Eats drivers in their grievances against the company but without any formal trade union accreditation (Uber Canada, 2022). Consequently, the United Food and Commercial Workers abandoned their request for formal accreditation before the OLRB. Henceforth, this settlement has been panned by large parts of the Canadian labour movement as a less-than-desirable third-tier option in comparison to the securities that formal accreditation would otherwise provide.
In the meantime, policy initiatives similar to the one spearheaded by the European Commission remain few and far between in Canada. Neither social democratic nor liberal governments at either jurisdictional level (provincial or federal) have proposed a significant policy to protect platform workers, especially those in high-risk/low-pay working conditions. Currently, the three provinces with the largest shares of platform workers in Canada-Alberta, Ontario, and Quebec-are all led by conservative governments. We can state, with certainty, that no imminent improvements at the policy level are in sight. In fact, the Ontario government recently passed a law enshrining independent contractor status for digital platform workers. This law only allows for very limited protections, such as respecting the minimum wage level and permitting individual dispute resolution (Government of Ontario, 2022).
As this section has shown, a concerted effort by established labour market actors such as trade unions, coupled with the mimicking effects and institutional disruptions within a federal structure, seems to nurture a conjuncture pushing governments towards policy experimentation. Much to the disadvantage of the workers, platform-based employment in Canada has not yet benefited from such a political conjuncture. With that in mind, let us now turn our attention to some examples from several EU member states.

Select Initiatives in EU Member States
There has been extensive research concerning experimentation with legislative frameworks offering protec-tions to app-based workers in Europe. While there are too many simultaneous developments to provide an exhaustive overview, the challenges of misclassification seem quite similar to those in Canada. Bennaars and Boot (2019) describe the contradictory rulings of Dutch labour courts on the classification of platform workers there, for example, opposing workers on two different market areas and platforms, such as food delivery (Deliveroo) and tourism (Booking).
Belgium briefly introduced its own version of a third status. The "De Croo law" (Verwilghen & Ghislain, 2020) principally aimed at clarifying the gig workers' status as small entrepreneurs under its tax law, while maintaining some elements of social protection as employees if they earned more than a certain monthly amount. However, after the noncompliance of several digital platforms, the law was partially revoked and the country reverted to more general protections reserved for those classified as workers (Raucent, 2022). It has, however, recently adopted a new law on platform work which presumes salaried employee status for platform-based work (FPS Employment, Labour and Social Dialogue, 2023).
In Denmark, trade unions succeeded in bringing at least two digital labour platforms (the cleaning staff platform Hilfr and the food delivery multinational Just Eat) into the prominent Scandinavian framework of voluntary collective agreements (Ilsoe, 2020;Scheele, 2021). As stated, these collective agreements are voluntary and not generally applicable by law. As a sectoral application of such agreements requires either the involvement of an employers' association or other major employers to sign up individually, there is now an active campaign to welcome the homegrown Nordic platform Wolt into the fold. If completed, this would leave the subsidiaries of the German multinational Delivery Heros, as well as Uber Eats, as the sole remaining major holdouts.
Recently, France has also legislated to reclassify app-based transport workers in the form of "dual status." This provides extensive individual and collective rights to workers while maintaining them as "independent contractors" (Ordonnance du 6 avril 2022, 2022). It remains to be seen, however, whether the enforcement of these rights will be effective. Compounding the challenges in that country, many workers are subletting their app accounts. This facilitates a much-needed source of income for undocumented migrant workers there (Gomes & Isidro, 2020). As France's version of a dual status continues to allow for subcontracting, we must now address the regularisation of highly vulnerable undocumented workers in the sector.
Based on a tripartite agreement between the government, two main employers' confederations, and two major trade unions, Spain promulgated the Rider's Law in May 2021 (Eurofound, 2021). It requires food delivery workers to be classified as salaried employees and digital labour platforms to disclose information about their algorithmic work organisation (e.g., payment schemes and schedules) to their employees. In adopting the law, the government and the social partners responded to a landmark Spanish Supreme Court ruling in September 2020. More recently, the Norwegian government (with the support of Norway's main trade union confederation) also proposed legislative action in the platform economy. Such a change would mandate employee status for digital platform workers (LO Norge, 2023).
Even in a country with dual labour relations systems (trade unions and statutory works councils) like Germany, recent strikes at the delivery platform Gorillas (a generalist, not specialising in food delivery) are also linked to problems of worker misclassification (Landesarbeitsgericht Berlin-Brandenburg, 2021). Lacking protections against dismissal in the case of collective action, the situation there quickly escalated to the labour courts. These courts decided that platform workers are correctly classified as independent contractors and thus may not engage in collective action. However, the federal labour court (Bundesarbeitsgericht) has simultaneously defined some criteria to determine salaried employee status for workers-similar to the Commission's proposal (Gramano & Stolzenberg, 2021).
Thus the European situation is comparable to its Canadian counterpart: Traditional labour relations actors are attempting to invoke traditional labour law, but the reality of platform-based work still largely escapes this route (with limited gains in Denmark). Despite legislators' timid attempts to start regulating the sector (e.g., Spain), enforcement dilemmas and collateral damages remain plentiful. Other legislators (e.g., in Belgium and France) have toyed with third-way or dual solutions. In the absence of a concerted campaign by labour market actors, such as national and European trade unions, it is unclear whether the momentum created by certain court rulings and the Commission's initiative is sustainable.

The European Commission to the Rescue? Analysing the Current EU Directive
It is in this volatile national-level policy context, described in the previous section, that litigation regarding the employment status of platform workers has risen sharply within the EU. Over 100 court decisions and 15 administrative decisions have been handed down since 2021's close (European Commission, 2021b). A comprehensive review of these decisions concluded with mixed results (Hießl, 2022): While British, Dutch, German, and Nordic rulings still maintain barriers between app-based work and salaried employee status, most decisions by national courts agreed to reclassify platform workers as salaried employees on a case-by-case basis. The European Court of Justice found the UK classification of platform workers to be compatible with EU law, as long as the workers remained independent contractors (B v. Yodel Delivery Network Ltd., 2020). To avoid a multitude of incoherent legal tests and the resulting patchwork of labour standards, with signifi-cant risks of so-called Delaware effects (Cary, 1974), the European Commission decided it was time to intervene.
To "support and complement the activities of the member states" (Consolidated version of the Treaty on the Functioning of the European Union, 2008), it drafted a directive to impose a uniform presumption of salaried employee status, applying to many of those working for digital labour platforms. If certain conditions were met, the presumption and subsequent reclassification would then automatically give the employees their rights under national legislation-including minimum wage (where it exists), working time and health protections and lastly unemployment and sickness benefits. All of the aforementioned would be determined according to where the work is performed, and not necessarily the home country of the digital platform-a principle previously enshrined in the revised posting-of-workers directive (Directive of the European Parliament and of the Council of 28 June 2018, 2018). Collective bargaining for autonomous workers is addressed in a separate set of Commission guidelines (Communication from the Commission, 2022) with national rules currently varying significantly (Fulton, 2018) between member states.
The initial criteria to warrant a presumption for reclassification were: 1. Unilateral wage setting; 2. Supervision, discipline, and algorithmic surveillance of performance; 3. Penalties for refusing shifts and/or prohibition of subcontracting; 4. Specific code of conduct for customer service and/or branded equipment and clothing; 5. Prohibition of "multi-apping" (working for other applications or as a truly independent contractor).
If two of these criteria were met, the directive would force member states to create a "presumption" of salaried employee status (European Commission, 2021a). The digital labour platform would then have to rebut the presumption in the labour courts, effectively inverting the burden of proof. However, the European Parliament, after discussing the draft directive (January 2023), did away with the five criteria completely and instead opted for a general presumption of employee status for platform workers. The European Council (June 2023) then reverted to a less directly-applicable presumption, leaving much of the burden of proof to the platform worker and their representatives. Once again, labour market actors will need to intervene in court if they seek to represent platform workers (entailing stark disincentives to organise workers from the start).
Whatever final form the EU directive will take, the original starting point (the initial five criteria) had a striking resemblance to the legal tests that Canadian jurisprudence developed to invalidate independent contractor status (either to reclassify workers as salaried employees, or at least to declare them as "dependent" contractors). Let us, therefore, contrast them in detail.
Canadian jurisprudence, both under Quebec civil law and according to the Anglo-Canadian common law tradition, also considers five criteria to determine whether a worker is an independent contractor (Gagnon, 2013): 1. Subordination, including surveillance, imposition of work schedules, and tasks, as well as creating economic dependency (by setting fee schedules unilaterally and prohibiting work for third parties); 2. Ownership of equipment and freedom to choose the geographical workplace; 3. The possibility to earn a share in the profits or risk of incurring losses; 4. The prohibition of subcontracting; 5. The integration into the employer's staffing structure and organisation of work.
Contrary to the EU criteria, they are not applied mathematically (two out of five). They are rather interpreted holistically and respecting a certain hierarchy, as demonstrated by the previously-discussed Foodora case. That hierarchy notably focuses on questions of subordination and economic dependency, while relegating the other criteria to a lower level. One must also keep in mind that the Canadian criteria were developed for all forms of independent contractors and that they were designed before the massive growth of digital labour platforms. Hence, the EU criteria are (of course) much more applicable to the reality of the platform economy. However, as Table 1 shows, they are still broadly comparable. It also becomes clear that the category of "subordination" and elements of "economic dependency" are understood to be the determining factors in Canada, while the other four criteria are supple-mentary. In the case of the EU directive regarding platform workers, this emphasis is also reflected by grouping three out of the five criteria regarding economic subordination (Table 1).

Discussion: Explaining Legislative Action on Platform Work (or Lack Thereof)
One attempt at reading the EU initiative might be a straightforward "spillover" argument (Haas, 1958), whereby the EU was pushed towards legislation. This momentum was provided by policymakers, labour courts, employers, and trade unions who were all making contradictory claims about the employment status of platform workers. According to this argument, maintaining coherence in the single market required the Commission's legislative intervention. A supplementary explanation stemming from sociological institutionalism lies in the isomorphism of intertwined jurisdictions (Sisson, 2007), whereby labour policy initiated in one jurisdiction may sway other member states or higherlevel governments into pursuing their own similar initiatives through mimicry. A more critical approach might view "social Europe" (Pochet, 2019) as serving as an antidote to successive EU crises. Especially the activism of the European Parliament, drastically broadening and deepening the proposed directive on platform work, seems to indicate a new urgency to occupy policy space, an area left relatively untouched by national legislators.
Compared to the EU, the Canadian experience provides none of the aforementioned conditions. The platform economy does not create significant spillover nor do mimicry effects among Canadian provinces (even less so within the minimal scope of federal labour legislation). Additionally, nor does the federal government see any political gains (or need for additional legitimacy) in The requirement to earn at least minimum wage in the country where the work is performed-not part of the original criteria, albeit one of the implied objectives The possibility to earn a share in the profits or risk of incurring losses The integration into the employer's staffing structure and organisation of work providing leadership on the issue. This experience contrasts somewhat with other recent policy initiatives by the Canadian government, such as the adoption of a proactive pay equity law (mimicry of previous Quebec legislation), the introduction of paid sick days (responding to the pandemic crisis), and even the revamping of a federal minimum wage as well as draft legislation for a more union-friendly labour code (both in response to political campaigns). One could thus argue that the federal government has now spent its political capital in the field of social policy, thereby making action on platform work highly unlikely. Why is it that millions of platform economy workers do not stimulate similar political activism, either at the behest of provincial governments or at the federal level? Everything points to the interaction of actor-based initiatives creatively mobilising existing legislation, mimicking limited decentralised initiatives elsewhere and creating urgency at the top level to regain the political initiative. While Canada does not seem to provide such a helpful conjuncture now, the EU just might.
Labour market actors often remain stuck in the traditional framework of labour law and in their very own "repertoires of action" (Tilly, 2006). They then neglect newcomers and outsiders, even when a sector is growing as rapidly as the platform economy. However, by changing the narrative about platform workers (such as overcoming the singularly judicial debate about employment status) and enlisting this dynamic workforce in creative political campaigns (Fulton, 2018), trade unions might yet become "strategic actors" (Hyman, 2007) in the sector. The precedent of the 2018 European Riders' Assembly in Brussels (Dufresne, 2019) and the nascent Gig Workers United in Canada (Canadian Union of Postal Workers, 2021) might mark the beginning of such an overarching social movement for platform workers.
Despite our fieldwork and analysis being limited to the food-delivery and transportation subsectors, most of the challenges concerning the legal status of platform workers and the blatant lack of congruent actorbased initiatives apply to many workers in the evergrowing gig economy. The range of professionalisation amongst these workers is also quite broad. On the lower end, one can find goods deliverers, homecare workers, administrative assistants, cleaning personnel, and maintenance workers. On the higher end, one can find programmers, stringers in journalism, graphic designers, and translators. Most of these workers face similar challenges of "dual" or "third" statuses under labour law. Many of them will satisfy legal tests on subordination and economic dependency. The overwhelming majority of them are currently unrepresented, not even courted, by labour-market actors such as trade unions.

Conclusions and the Way Ahead
Returning to our initial research subject once again, in Canada, most platform workers work under provincial labour law rather than under federal jurisdiction. Even though the mimicking effects of federalism can (and do) impact legislation eventually adopted by the provinces, progress has been limited. As addressed earlier, the current political orientation of the three major provincial governments (Alberta, Ontario, and Quebec) makes progress at that level unlikely. The federal political climate, despite a liberal minority government supported by a social democratic party, is unlikely to foster the necessary momentum either.
So what can Canada learn from the EU? If a concerted social movement like the one calling for federal leadership on a $15 minimum wage, spillover, or mimicry mechanisms are off the table for now, what is the next step? As we have seen in the European cases (with individual national governments, various local labour courts, and administrative bodies adopting reclassification), approaches driven by labour market actors alone are unlikely to succeed. At best, they might simply create a patchwork of employment statuses and highly uneven social rights. In that context, it is laudable that some Canadian trade unions (such as the Canadian Union of Postal Workers) are cooperating with platform workers to create a sensible litigation strategy. But a concerted, Canada-wide campaign in favour of regulating platform workers' rights (like the EU directive), is still far from being on the horizon.
Even once critical mass is reached and legislative action is initiated at the top level, the road to implementation and overcoming resistance to subsidiarity in multi-level governance systems is long. In the EU, Council and member states may have already undercut the proposed directive before it was able to create a level playing field for the platform economy. Trade unions will need to intervene to preserve the existing progress.
Based on our analysis, the importance of mutually reinforcing mechanisms between labour market actors and policymakers is paramount. These mechanisms further underscore the theoretical arguments calling for meaningful institutional experimentation within multi-level governance systems. Actor-based initiatives alone cannot explain progress in this field, and neither can legislative spillover and mimicking effects-upwards or sideways.
Scholars may disagree on the precise role of labour market actors, but (as this article has attempted to demonstrate) these actors do play important roles. They both advance bottom-up processes and promote political urgency at the relevant institutional levels, be it through a timely litigation strategy or concerted political campaigns. The inclusion of platform workers within established labour unions on both sides of the Atlantic is thus an important precondition for other social actors (e.g., employers, labour courts, and policymakers) to embark on their own pathway towards meaningful regulation.
So, what is the future for platform workers? The jury is still out on the EU directive and its implementation. In Canada, limited gains by trade unions have quickly been swallowed up and replaced by employers' initiatives. In the meantime, the expansion of employment in the platform economy is unrelenting. The current shortage of labour supply in many Western countries may tilt the bargaining power in favour of platform workers, but it is unclear whether their actions can be concerted enough to force meaningful and enduring policy change. All the while, the grave dangers of (some) platform work will continue to push vulnerable workers to "complete the delivery" while federalist policymaking plays catch up.