The employment status of food delivery riders in Europe and the UK: Self-employed or worker?

Online platforms are revolutionizing our daily lives in an attempt to make it easier by offering innovative services. They also have introduced radical new business models which provide a new type of flexible working, facilitating employment. While platforms are revolutionary vehicles, they also denied workers status, resulting in food delivery riders facing precarious working conditions. The current regulatory framework is underdeveloped and unable to guarantee basic social rights to platform workers, except for Spain. At the same time, delivery workers are fighting to get some form of recognition and protection. Consequently, courts have been increasingly requested to determine the riders’ legal status. However, courts are struggling in characterizing those employment relationships resulting in disparities. For instance, the Cour de Cassation in France has established that an employer-employee relationship existed while the UK High Court denied worker status to Deliveroo riders. This lack of harmonization and different rulings could result in the application of EU rules in some countries but not others. It might, therefore, be time for the EU to start recognizing and regulating these jobs to offer better worker protections.


Introduction
Online platforms are revolutionizing our daily lives in an attempt to make it easier by offering innovative services. This revolution came at the cost of increased pressures on the existing regulatory framework. It soon became clear that the current regulatory framework is unable to guarantee basic social rights to platform workers, except for Spain. 1 For obvious reasons, delivery platforms have tried to deny any employment relationship.
The heavy reliance on self-employment status brings obvious risks due to the inadequacy of the current legal framework. Indeed, most worker protections, such as minimum wages and social insurance, are directly linked to the employment relationship. Food delivery riders 2 often struggle with unregulated working conditions while being obliged to comply with the technology platforms' terms of use and increasing competition that these companies bring. This struggle results in precarious conditions for solo self-employed (i.e. self-employed without employees) who are highly dependent on the platforms. 3 The high dependency on platforms means that such platforms hold excessive power, leading to social dumping. At the same time, delivery workers are fighting to get some form of recognition and protection. Consequently, courts have been increasingly requested to determine the riders' legal status. However, courts are struggling in characterizing those employment relationships resulting in disparities. For instance, the Cour de Cassation in France has established that an employer-employee relationship existed while the UK High Court denied worker status to Deliveroo riders. 4 To complicate an already complex situation even more, some courts have differentiated Uber drivers from delivery riders on the simple argument that Uber drivers cannot outsource their work. While this is undeniable, such differentiation does not take into consideration the fact that Uber Eats is based on the Uber model, meaning that the terms and conditions are pretty similar. This differentiation could partially be explained by the great misconceptions regarding the platform economy, which slows any adaption of the existing laws.
The COVID pandemic has highlighted the importance of the service provided by those food delivery workers. In many countries, they have allowed restaurants to continue working during lockdowns. Although they provided valuable services, the status and remunerations of the food delivery workers have not changed. It might, therefore, be time for the EU to start recognizing and regulating these jobs to offer better worker protections. This is especially true as the divergences in rulings are endangering the existing harmonization. This paper will, therefore, evaluate the consequences of the divergence of judgments at the national level and the laissez-faire at the EU level and will argue that riders should be granted some form of employment benefits as considering riders as full employees will hinder the flexibility that defines platform work. Section 3 analyses the divergent judgements at the national level whereas section 4 is dedicated to analysing the decisions in the UK. Section 5 evaluates the European regulatory response and highlights its deficiencies. Finally, Section 6 discusses the Court of Justice of the European Union (CJEU) opinion on the matter.

Why is it a problem?
Online platforms have introduced radical new business models which provide a new type of flexible working, facilitating employment. 5 While they are revolutionary vehicles empowering both customers and workers, they are also associated with the rise of 'cybertariat', which benefits owners over workers. 6 Indeed, platforms are negating the employment relationship between themselves and the workers. As Drahokoupil and Piasna noted, 'The reorganization into self-employment of activities that traditionally offered opportunities for employment represents the key transformative marketmaking potential of platforms.' 7 The riders' high dependency on platforms means that such platforms hold excessive power, leading to social dumping. As Prassl and Risak argued, 'Individual platforms' terms and conditions vary from country to country according to local conditions, whilst always pursuing identical aims: the denial of worker status.' 8 Consequently, the lack of adequate regulation results in employment's precariousness and risk-taking while benefiting major companies. 9 Finally, platforms have intensified already existing problems such as outsourcing. 10 The current situation allows major companies to take advantages of the system and avoid paying employment taxes while exercising an often-significant level of control over the workers. This is especially worrisome that most riders rely on delivery works as their primary income source, such as students, and not just as a top-up. As Mason noted, 'Trade unions estimate that around half a million of those are bogus and are really working for a single employer, using the status to collude with that employer to pay less tax. But for them, the traditional trade-off of selfemploymentlower tax and national insurance in return for fewer statutory benefitsis not always a choice'. 11 Similarly, Adams-Prassl and Risak argued that 'individual platforms' terms and conditions vary from country to country according to local conditions, whilst always pursuing identical aims: the denial of worker status'. 12 As Drahokoupil and Piasna noted, 'The usage of these terms seems [collaborative and sharing economy] to reflect efforts to cast these new phenomena as something inherently positive, which is not helpful to keeping the policy debate evidence-based and free of any pre-conceived biases. The notion of the sharing economy can even be linked to the lobbying efforts of the major platforms.' 13 Sharing economy is still regarded as addressing market failures with less emphasis on the redistribution of risks and costs. In fact, the market failure created by the gig economy, highlighted by the precariousness of the worker status, is yet to be analysed.
By allowing food delivery workers to be regarded as self-employed rather than workers or employees, States indirectly participate in the endangerment of riders' lives. Indeed, it is not uncommon to read newspapers articles about delivery workers who have been severely injured or killed while delivering food for a platform. Workers are encouraged to reach targets which can push them to commit some delicts. 14 Moreover, the navigation system of Uber Eats is programmed for a car and not a bike, leading to some dangerous situations with riders being sent on a highway. 15 There is, therefore, a gap between reality and promises made by the platforms which require state intervention.

Divergent practices across member states
Countries are divided regarding delivery riders' status; in some countries, there is clear case law establishing that delivery workers are employees. In other countries, their self-employed status has been maintained.

A. France
Although the French legislator adapted Articles L.7341-1 and L.7341-6 in 2016 to provide some minimum social rights to workers, these rights have been barely respected, especially the work insurance coverage. However, the status of the worker was not clarified. As a result, employment tribunals have been increasingly asked to answer whether independent workers can really be considered independent contractors or if, instead, they are, in fact, employees.
The first appeal judgment on that topic was the 2017 Court of Appeal judgment in favour of Deliveroo. 16 The Court of Appeal ruled that there were not enough evidences of a subordination tie to the company due to the freedom the riders enjoyed.
In 2018, the French Supreme Court ruled that a Take Eat Easy worker was an employee. 17 The Supreme Court overturned a judgment of the Court of Appeal that found no employment relationship between Take Eat Easy workers and the platform. The Supreme Court disagreed and found that the contract violated Article L8221-6 of the French Labour Code because of the company's direct control over the workers. Moreover, the fact that the platform was able to punish and reward workers was another element demonstrating an employment relationship. The Court disregarded the arguments that there was no exclusivity relationship and that the worker was free to organize his or her time. The geo-localization system allows the platforms to know where the workers are, and the counting of kilometres proves a form of control in the execution of the services. This judgment applies to other platforms. However, looking at Uber Eats' website, the company still advertises new jobs as self-employed.
The 2018 Supreme Court decision was quickly followed by a decision in the Paris Court of Appeal 18 which ruled that an Uber driver operated under an employment contract, which the Supreme Court confirmed in March 2020. 19 In 2020, the Industrial Tribunal of Paris stated that the 'services agreement entered into with Deliveroo France shall be seen as an employment agreement'. 20 This judgment is in line with the Take Eat Easy judgment using the same reasoning, namely the GPS system allowing real-time tracking and the power to inflict sanctions. Deliveroo France was, therefore, condemn to pay termination indemnities and damages for unfair dismissal. As Teixeira and Martel noted, 'the judges went even further. They also judged that Deliveroo France was guilty of concealed employment, considering that the Company had deliberately circumvented the formalities associated with the hiring of employees, the payment of social security contributions and the remittance of payslips.' 21 However, soon after these rulings in favour of the riders, various other decisions have taken the platforms' side, with courts refusing the possibility of an automatic requalification of those contracts. For instance, the Paris Court of Appeal held that there was no permanent legal subordination link in the Tok Tok Tok cases. 22 A similar reasoning is found in the judgment in favour of Uber from the Lyon Court of Appeal. 23 In April 2021, the Paris Court of Appeal confirmed the first instance judgment and its previous position on Deliveroo by rejecting a delivery rider's request to have his contract reclassified as an employment contract. 24 The rider claimed that the requalification of his contract should be quasi-automatic pursuant to two rulings of the Court of Cassation. After conducting an in-depth analysis, the court dismissed any permanent legal subordination link and therefore the existence of an employment contract.
While those decisions seem to directly conflict with the 2018 French Supreme Court ruling, they are consistent with the European Court of Justice in Yodel. 25 Consequently, according to those latest judgments, it can be concluded that a platform worker is not an employee. In order to requalify an employment contract, the worker must provide evidence of the existence of a permanent legal subordination. Those evidences cannot be based on general elements that are inherent to any commercial relationship with a digital platform but instead must be specifically applicable to the worker's personal situation. These judgments raise even more concerns with regard to the level of protection afforded to riders.
Only the enactment of a new law would crystalize the rulings in favour of riders, which seems unlikely. In 2020 a bill to stop exploitation and offer better recognition was rejected by the Senate. 26 The platform workers representatives were hoping for a positive outcome in light of the COVID pandemic, unfortunately rejected. 27 The status of riders could have been even more weakened without the Conseil constitutionnel decision regarding the validity of Article 44 of the brand new Loi d'Orientation des Mobilités (LOM). 28 Indeed, that specific article would have greatly restricted the possibilities of a judge to requalify the relationship between platforms and riders as an employment contract. Although this new law provides additional protection, 29 the workers are still not adequately protected.

B. Belgium
Belgium is maybe one of the most interesting countries to analyse. At first, non-standard workers received little regulatory attention in Belgium. Consequently, several initiatives have been developed to circumvent the problems by relying on labour market intermediaries (LMI). 30 One of the most visible initiatives is the Société Mutuelle pour ARTistes (SMart), created in 1998 to address the absence of appropriate employment solutions for artists. 31 The system was then extended to project-based workers. SMart developed two main tools to provide more security to members: first, employment contracts and second, activity management. The employment contract aims at guaranteeing the members' pay while reducing their paperwork. As an LMI, 'SMart invoices the contractors on behalf of its members and returns the money back to the members as salary in the framework of an employment contract'. 32 This system also guarantees an employee status for the time of the work.
In 2013 a Belgian startup, 'Take Eat Easy', started operating. Due to the nature of the job, some riders started using SMart, but the number was only marginal. 33 However, it is only with the arrival of Deliveroo in the Belgian market in 2016 that the use of SMart employment contracts increased. Indeed, the SMart system was (and still is) an excellent alternative to the self-employed status. In fact, SMart developed a joint protocol with Deliveroo and Take Eat Easy in 2016 to improve riders' 26. Sénat, 'Proposition de loi relative au statut des travailleurs des plateformes numériques' (2019), www.senat.fr/dossierlegislatif/ppl18-717.html. 27. Fabien Recker, 'Plateformes numériques: une proposition de loi pour protéger les travailleurs', Public Senat (2020), www.
publicsenat.fr/article/parlementaire/plateformes-numeriques-une-proposition-de-loi-pour-proteger-les-travailleurs. 28. Décision no. 2019-794 DC du 20 décembre 2019. 29. Better information on the fees, the right to refuse to do a delivery or a transportation, the right to choose the work schedule, etc. working conditions and pay. Indeed, they discovered that riders were obliged to declare fewer hours than they effectively worked to comply with the legal minimum pay. 34 The use of SMart employment contracts is not only significant for workers, but it also provides reliable accident rates. These statistics have helped established that riders had 10 times more work-related accidents than the national average. The SMart system might be one of the most adequate solutions as it offers flexibility to those who want it while offering protection to others. Although a system like the SMart system in Belgium and its Dutch counterpart seems a good option, it would be quite challenging to implement it at EU level. Meanwhile, two unfavourable measures were adopted: the so-called De Croo measures/law 35 and the creation of a self-employed student status. 36 Uber Eats, which entered the Belgian market at the end of 2016, only relied on these measures and never cooperated with SMart. As a result, in October 2017, Deliveroo announced its intent to only 'hire' self-employed riders and 'asked' its riders to leave the SMart system, resulting in a decrease in labour conditions. 37 According to SMart, under their system, riders earned 11€/h plus 2€ per order as self-employed or 9.49€ as SMart employees plus 0.12€/hour for their phone and 50% of any repair to their bikes. 38 Besides, they were insured and had the certainty of being paid at least three hours per day worked. Currently, self-employed are paid 7.25€ per ride, while students and workers under De Croo law are paid 5€ per order. These laws have been quickly criticized due to the possible deprofessionalization they would create but also their discriminatory nature. 39 in 2020, the Belgian Constitutional Court annulled the three pillars in the 2018 law. 40 In Belgian law, as is the case in most countries, it is possible to requalify a labour contract. However, this mechanism is only applicable on a case-by-case basis. 41 Although there is currently no decision from the Belgian Supreme Court, a lawsuit was filed by two Deliveroo riders. In March 2018, the Commission administrative de règlement de la relation de travail (CRT), an administrative body, concluded that the nature of the work could not be qualified as self-employed work. The decision lists similar reasons as the French Supreme Court; parties' intention, freedom to organize work and working time and hierarchical control. 42 The existence of a link of subordination is the criterion that fundamentally distinguishes an employment relationship from a self-employed relationship. 43 The decision was mainly based on Article 337, section 1, 3°of the Belgian Labour Relations Act of 27 December 2006, as amended in 2013, which relates to the transport of goods and persons for a third party. While CRT rulings are not binding, they are highly relevant. Deliveroo contested this ruling in front of the Brussels Labour Tribunal by bringing a claim against the two riders and the Belgian State. 44 In 2019, the Tribunal ruled that the CRT decision was invalid because of an ongoing investigation. 45 In 2020, the Office National de Sécurité Sociale (ONSS) started court proceedings against Deliveroo for unpaid social contributions. 46 This proceeding is based on a two-year investigation which concluded that riders are employees. The procedure is still ongoing, with the first hearing is scheduled for October 2021. If the tribunal agrees with the conclusion of the investigation, it could bring significant changes for riders. Additionally, the fact that the ONSS has filed a lawsuit in front of the Brussels Labour Tribunal is fascinating. Indeed, unlike most lawsuits started in other EU countries, in this case, it is not riders that wanted their status to be recognized by a governmental entity that is against the reduction of riders' rights. While party autonomy is the predominant legal principle, Belgian courts might follow the French example and look at the contract's factual performance rather than just its classification.
In October 2020, the CRT concluded that a subordination link existed between Uber and its drivers. 47 The CRT ruled that the evidence was incompatible with self-employed status.
Finally, there have been several rulings by the Belgian Administrative Commission for the Determination of the Employment Relationship, which concluded that the Deliveroo riders should be considered as employees. 48 While these rulings are not binding, they are highly relevant. 49 In 2021, a bill was introduced to grant riders an employee status and get rid of fake selfemployment contracts. 50 This bill follows the example of the brand-new Spanish riders' law.

C. Spain
As Todolí Signes rightly argued, 'Spain is one of the countries with the highest levels of judicialization of the dispute over the classification of platform working.' 51 Indeed, there have been dozens of rulings in recent years. While most of those rulings consider riders as workers, some have categorized riders as self-employed. 52  some 500 riders were, in fact, Deliveroo workers. 53 Additionally, Deliveroo was ordered to pay 1.3 million euros in unpaid social contributions following a court ruling by the 24th Social tribunal in Barcelona, which established that 748 Deliveroo's riders are false self-employed. 54 The issue is now settled after the Spanish Supreme Court ruling, which unanimously declared that riders are workers, putting an end to the conflicting judgments. 55 Interestingly, the case given rise to this ruling was first decided by the Social Court of Madrid, which found the rider to be genuinely self-employed. This ruling was confirmed by the High Court of Justice in Madrid and overturned by the Supreme Court. One of the main elements that the Court factored in was the platform's central role. The Spanish Supreme Court referred to Yodel but refused to request a preliminary ruling. 56 More recently, Spain regulated this matter by enacting a specific legislation; the so-called Rider Law. Spain will be the first EU country to give gig economy workers an explicit salaried employee status. Interestingly, to make sure Parliament will not reject the text, the government decided to pass it as a legislative decree. The legislation's main advantage is probably the certainty of the number of hours riders have to work. This will avoid 10-hour shifts which are common in other EU countries. While the Rider Law was not yet published at the time of writing, it has already been argued that platforms could still hire self-employed workers and go around the legal presumption found in the Rider Law. 57 Although not yet published, the law has already been widely criticized, especially by platforms themselves. 58 It can be wondered whether the Rider Law is the most adequate manner to solve the problem. On the one hand, it will ensure that riders are protected as they should. On the other hand, it will hinder flexibility; some riders need flexible hours/days for personal reasons, such as caring for someone. There have been reports of chronically ill riders who favoured the current system as it allowed them to work when they felt able to but did not require them to call in sick. Finally, the potentially harmful effect of the law will have to be assessed in the medium run.
This law could influence other countries, especially Austria, to enact new measures. Indeed, 'a renowned Austrian professor submitted a draft law that tackles the issues related to platform work and offers a rebuttable presumption for the existence of an employment contract. If such a law were passed, this would result in a reversal of the burden of proof by which the platform company would have to rebut the presumption of an employee's status.' 59 Although there was no political will to enact such law at the time of submission, the introduction of the Rider Law in Spain could boost the Austrian proposal.

D. Other member states
Like most countries, Italian jurisprudence was, up until 2021, divided on food delivery riders' status. Interestingly, the first cases were against the granting of worker status to riders. 60 The reason that the two courts reached the same outcome was a narrow interpretation of the notion of subordination found in Article 2094 of the Italian Civil Code. 61 However, a change in the labour minister and the COVID pandemic has led to drastic changes; In 2021, the Italian Labour Inspectorate and Milan prosecutors ordered food delivery platforms to pay €733 million in fines and to fully hire the riders. 62 This decision goes against various earlier judgments which established that riders were self-employed. 63 Interestingly, while the Turin Court of Appeal confirmed the first instance judgment, it also ruled that riders cannot be considered fully self-employed and establishing that Foodora riders belong to a third category which is neither self-employed nor subordinated employees. 64 The Italian Supreme Court confirmed this ruling by concluding that riders are not employees but should be granted some of the employees' basic rights.
At first, riders in The Netherlands were protected by employment contracts until 2017. 65 Since then, conflicting verdicts were issued in The Netherlands. 66 Some of these conflicting rulings can be explained by the switch from employment contracts to self-employed contracts. For instance, the first Deliveroo ruling was based on an employment contract, while the second was with a 'selfemployed' rider. 67 More recently, the Amsterdam Appeal court ruled that Deliveroo riders are employees based on the existence of a relationship of authority. 68 The recent ruling by the German Federal Employment Court has made clear that crowdworkers are employees. 69 This ruling, therefore, suggests that riders will also be classified as employees. This judgment departed from the position taken at first and second instance. 70 The Court reached that decision because of the existing subordination, and the contractual freedom of crowdworkers is restricted. 71 Interestingly, 'the Court based its recognition of the crowdworker's personal dependence on the platform operator (as the employer) primarily on the psychological effect of the incentivized system, which was stated to have induced the crowdworker to continuously carry out work for the platform operator.' 72 However, this ruling does not mean that food delivery riders will be de facto employees. The Federal Ministry of Labour and Social Affairs has recently published the 'Fair Work in the Platform Economy' report, which aims to improve the legal position of platform workers. The German government also wants to shift the burden of proof to make it easier for platform workers to be granted employees status and rights. Nevertheless, compared to other countries, there has been no specific cases decided regarding food delivery riders. Finally, Deliveroo withdrew from the market in 2019. 73 In Poland, there is no case law on the topic despite the ongoing debates in the media. The Polish parliament has not taken any action. So far, riders are self-employed. 74 Similar situations are found in other Eastern European Countries. Interestingly, there have been no case laws or attempt to regulate in the Nordic countries. 75 This is surprising because the platform economy 'challenge[s] the Nordic model of work and welfare'. 76 While no case law or bills are being discussed, the Finnish Labour Council issued two opinions concluding that riders are employees who fall within the scope of the Working Hours Act. 77 Like in Belgium, these decisions are not binding, although highly relevant. 78 In 2019, Foodora riders in Norway concluded a collective bargaining agreement (CBA). 79 Therefore, it could be assumed that riders might be granted employees status if their status was ever challenged in court.
Finally, there is no case law or regulatory framework in place in Austria, meaning that the general frameworks are applicable. 80 The classification of a delivery worker as an employee or selfemployed will depend on the 'usual criteria defining the dependent work of an employee'. 81 These criteria include dependency, behavioural control, disciplinary authority and operating resources. As noted above, pressure is slowly mounting, with a draft law submitted by a renowned professor. 82

UK: Are delivery workers contracts sham contracts?
In the UK, the rights provided by the law to an individual depend on his or her employment status. However, establishing the employment status of an individual is not always straightforward. Previously, the focus was on the written agreement of employment, which brought legal certainty. This approach was dangerous as the person drafting the contract is usually the employer. The contract, therefore, reflected the employer's presentation of the relationship, which in some cases was not accurate. Such heavy inequality in the bargaining power created a risk of 'sham' contracts.
Sham contracts give the impression that the employee is self-employed while, in reality, the person is an ordinary employee. In the wording of Lord Diplockwho first defined sham contracts in Snook v. London & West Riding Investment Ltd, they are contracts which are 'different from the actual legal rights and obligations (if any) which the parties intend to create'. 83 Consequently, a sham contract does not accurately reflect the de facto agreement. These contracts unlawfully deprive individuals of their statutory employment rights under the Employment Rights Act 1996, such as the right not to be unfairly dismissed, but also would place them in the position of a real self-employed, paying their tax and national insurance. 84 Sham contracts are, therefore, a misrepresentation of employment status and are prohibited.
Courts have ruled that the approach to sham contracts which apply to ordinary contracts should be altered in employment situations. 85 Consequently, the courts developed a test that focuses on the reality rather than the written documents, such as in Consistent Group Ltd v. Kalwak and others. 86 In this case, however, the Court of Appeal overturned the Employment Appeal Tribunal (EAT) decision, on the ground that the tribunal had not given sufficient reasons for its decision that the clause stating that there were no obligations between the parties was a sham. Interestingly, the test set in Consistent and the analysis made by the EAT was heavily relied on in Firthglow Ltd (Protectacoat) v. Szilagyi, where the principle was extended to cover self-employment contracts and noted that 'the test for a sham must be sensitive to context'. 87 It was held that the terms must accurately reflect the situation 'not only at the inception of the contract but as time goes by'. 88 Finally, the Court of Appeal and the Supreme Court's decisions in Autoclenz Ltd v. Belcher and others brought the changes intended by the employment tribunals in Consistent Group Ltd v. Kalwak.
In a more recent case, an employment contract labelling the 'contractor' as self-employed was defined as a sham and declared that the contract was designed to mask the actual relationship. 89 In this case, although the courier had the right to use a substitute, once she was logged on the company's tracker system, she was expected to accept jobs given by the dispatcher and be ready. However, the peculiarities of this case could explain the outcome. 90 Finally, a key decision on sham contracts and the gig economy is the Supreme Court ruling in 2018 in the Pimlico Plumbers case. 91 In this case, the Supreme Court upheld the Court of Appeal's decision, which found a plumber to be a worker. 83 The judgment sent a strong signal that 'simply labelling workers self-employed does not guarantee the corresponding legal status. The nature of the relationship and the degree of bargaining power and obligation between the parties is crucial in determining workers' rights.' 92 Unfortunately, this ruling does not lay down new principles but rested heavily on this case's facts, as the Court mentioned.
Interestingly, while there are many cases challenging sham contracts, including an Uber case, those cases do not apply to food delivery riders. Although delivery workers' contractual situations bear similarities with sham contracts, the High Court did not buy this argument. Instead, the Court ruled, in 2019, that Deliveroo riders were not workers but contractors, based on the fact that the riders were able to abandon a job or pass it on to another rider. 93 The fact that the platforms know where the riders are through the geo-localization system and kilometres were not regarded as establishing a worker status, as it did in France. This ruling was confirmed by the Court of Appeal in 2021. 94 In fact, courts have been reluctant to analyse how the contract was actually performed. In Tanton, for instance, Peter Gibson LJ ruled that the question related to the legal obligations rather than how the contract was actually performed. 95 Interestingly, the overemphasis on substitution was addressed in the Taylor Review in 2017, which recommended that the test for worker status should be more focused on control and less on personal service. 96 Unfortunately, this report did not impact the judiciary much.
While Deliveroo riders are not employees or workers for collective bargaining purposes, the Court of Appeal found that Uber drivers are workers and not self-employed when they switch on the app. These two cases are somewhat conflicting, as driving for Uber or riding for Uber Eats are involved in similar actions. The UK Supreme Court upheld this ruling in February 2021. 97 In fact, that judgment could appease this conflict. The Supreme Court judgment emphasizes five aspects of the findings made by the employment tribunal, which justified its conclusion that the claimants were workers; first: Uber sets the fare and drivers are not permitted to charge more than the fare calculated by the Uber app. Second, the contract terms on which drivers perform their services are imposed by Uber, and drivers have no say in them. Third, once a driver has logged onto the Uber app, the driver's choice about whether to accept requests for rides is constrained by Uber. One way in which this is done is by monitoring the driver's rate of acceptance (and cancellation) of trip requests and imposing what amounts to a penalty if too many trip requests are declined or cancelled by automatically logging the driver off the Uber app for ten minutes, thereby preventing the driver from working until allowed to log back on. Fourth, Uber also exercises significant control over the way in which drivers deliver their services. A fifth significant factor is that Uber restricts communications between passenger and driver to the minimum necessary to perform the particular trip and takes active steps to prevent drivers from establishing any relationship with a passenger capable of extending beyond an individual ride. 98 The same five aspects apply to food delivery riders; the riders do not set the fare, the platforms impose the terms, the acceptance rate is monitored, 99 platforms exercise significant control over the manner the service is provided and, finally, communications between the riders and clients are limited. According to these criteria, delivery riders could be workers and their contracts regarded as sham contracts. Unfortunately, the Supreme Court did not directly address whether the drivers were also employees.
Under English law, the status of worker is something of an in-between or hybrid status; they enjoy some key legal rights but are entitled to fewer statutory rights than employees. 100 Moreover, this status implies that the person personally performs the work in question. 101 This inability to subcontract work under the worker status is one of the main reasons for qualifying delivery riders as self-employed. 102 By recognizing delivery riders as workers, it could also put an end to the worker black market. Indeed, a UK investigation has uncovered that foreigners on a holiday visa are paying to use another rider's account for the time of their 'holidays' and then go back home without ever declaring the amount they earned. 103 While the granting of employee's status would bring significant benefits, it seems very improbable that it will be awarded to delivery riders any time soon. In fact, as demonstrated by the High Court judgment and the Central Arbitration Committee (CAC), a distinction is made between delivery and other gig economy jobs. 104 Even if a court grants delivery riders a worker's status, the situation might evolve in the same way as in the construction industry; After the government legislated in 2014, the workers were required to operate via 'umbrella companies'. 105 A glimpse of hope comes from the draft bill produced by the Work and Pensions Committee and Business, Energy and Industrial Strategy (BEIS) Committee designed to tackle the perceived exploitation of 'gig economy' workers. This draft bill is a response to Taylor's review of modern employment practices. 106 The proposal would bring change to primary and secondary legislation by introducing a new definition of self-employment but also by imposing a 'worker by default' model for businesses that use a substantial number of self-employed staff. The bill will also prevent companies from using false self-employment status for tax avoidance and cheap labour. The bill has, however, not yet been enacted.

European regulatory response and its deficiencies
At the EU level, platforms are in a state of nearly complete laissez-faire; the lack of any regulatory framework allows platform owners to set the rules. This lack of regulatory framework also results in workers getting their rights recognized only after a court ruling. While the Yodel judgment brings some clarifications, it does not offer a complete framework. This, in turn, jeopardizes any harmonization attempts due to the broad discretion left to national courts.
The European Commission's Communication on the Collaborative Economy could play a crucial role, as it aims to establish a regulatory framework for platform work. 107 It tried to define sharing economy. 108 The Commission also advises national regulatory authorities to recognize platforms' specificities while maintaining the protection afforded to employees. The Commission acknowledges the 'need to prevent the platforms from being a driver of excessive deregulation of employment'. 109 Indeed, the regulatory challenges are manifold and finding the adequate balance is not easy. 110 This Communication, however, does not address the specific challenges faced by delivery workers and platform workers in general. The major obstacle for platform workers is that the platforms are often not registered in the same country as where the work is carried out. For instance, Deliveroo is registered in the UK but also operates in France, Belgium, Italy, Spain and The Netherlands. Therefore, it is crucial to have an EU framework in place that recognizes delivery workers as employees.
Additionally, the European Committee of Social Rights, in its decision concerning the right to collective bargaining of self-employed workers, observed: 'the world of work is changing rapidly and fundamentally with a proliferation of contractual arrangements, often with the express aim of avoiding contracts of employment under labour law, of shifting risk from the labour engager to the labour provider'. 111 Another promising proposal for platform workers is the Commission's recent proposals regarding collective bargaining for the self-employed with a view to the possible adoption of a Council Regulation. 112 The Commission felt that the new form of work introduced by platforms and digitalization introduced uncertainty regarding access to collective bargaining. While collective bargaining is a powerful tool to improve working conditions, competition law provisionsespecially Article 101 TFEUare an obstacle for self-employed workers because they are considered undertakings. The proposal contains four options which differ based on the categories of worker. Option 1 applies to 'All solo self-employed workers who provide their own labour through digital labour platforms'. Option 2 applies to 'All solo self-employed workers who provide their own labour through digital labour platforms or to professional customers of a certain minimum size'. Option 3 includes 'All solo self-employed workers who provide their own labour through digital platforms or to professional customers of any size with the exception of regulated (and liberal) professions'. Finally, option 4 applies to 'All solo self-employed workers who provide their own labour through digital labour platforms or to professional customers of any size'. This proposal will not per se lead to changes in the status of food delivery workers. However, it will allow solo self-employed workers to access collective bargaining, resulting in significant changes. While it is not an ideal solution, the proposals bring hope that things might change.
The exclusions of gig economy workers from fundamental labour rights and the limitations regarding the exercise of collective rights give undue incentives to recur to non-standard works. The lack of regulation and clear position at EU level not only affects food delivery workers but also all non-standard workers, resulting in a form of social dumping. For instance, bogus selfemployment contracts have been flourishing in the aviation sectors for decades without much discussion. 113 By enacting rules at EU level in this specific field, it will send strong signals to other areas that recur to non-standard works. It is, therefore, time for the EU to start regulating those practices to avoid any infringement of workers' rights and help harmonize the situation in Europe, especially after the Yodel ruling.

Court of Justice of the European Union (CJEU) opinion on the matter
Unfortunately, no case relating to food delivery workers has yet been analysed by the Court of Justice of the European Union (CJEU). The Court has already given its opinion regarding Uber, although not on the employment status per se, and more recently on food delivery riders. However, in 2020, the Court provided some clarification on the definition of 'worker' status under EU law in a 'reasoned order'. 114 The case is fairly unusual as the reference for a preliminary hearing came shortly before the UK was due to leave the EU. The case involves a parcel delivery courier, 'B', who carries out his work exclusively for Yodel under a self-employment contract.
The CJEU was asked whether the right to engage subcontractors or 'substitutes' to perform all or any part of their work meant that an individual could not be regarded as a worker under the Working Time Directive. 115 The Court started by recalling that the term 'worker' has autonomous meaning under EU law, even if it is not defined in the Directive. 116 The Court ruled that national courts must: determine to what extent a person carries on activities under the direction of another, base that classification on objective criteria and make an overall assessment of all the circumstances of the case brought before it, having regard both to the nature of the activities concerned and the relationship of the parties involved. 117 Whilst recognizing that the final outcome of the case is a matter for the domestic court, the CJEU did express a firm view: 'In the light of all those factors, first, the independence of a courier, such as that at issue in the main proceedings, does not appear to be fictitious and, second, there does not appear, a priori, to be a relationship of subordination between him and his putative employer'. 118 Although it was the employment tribunal's job to determine the courier's employment status, the CJEU statement seems quite clear; B is not Yodel's worker.
The CJEU still provided indications, 'in order to give a useful answer to the referring court'. 119 First, the Court pointed to the substantial amount of latitude B had in relation to his employer. 120 Indeed, it was necessary to determine the consequences of that latitude on the independence of B or whether his independence was merely notional. 121 Another crucial element was whether B was in a subordination relationship with Yodel. 122 It was also significant that the limitations on B's right to provide a subcontractor or substitute were minimal. Essentially, the substitution could be anyone who had basic qualifications and skills for the job equivalent to B. 123 B had an absolute right to accept or reject tasks assigned to him alongside the fact that B had the right to provide his services to Yodel's direct competitors simultaneously. 124 Finally, while B had to deliver the services within particular timeslots that simply reflected the inherent nature of Yodel's business. 125 Considering all those factors, the Court held that: Directive 2003/88 must be interpreted as precluding a person engaged by his putative employer under a services agreement which stipulates that he is a self-employed independent contractor from being classified as a 'worker' for the purposes of that directive, where that person is afforded discretion: • to use subcontractors or substitutes to perform the service which he has undertaken to provide; • to accept or not accept the various tasks offered by his putative employer, or unilaterally set the maximum number of those tasks; • to provide his services to any third party, including direct competitors of the putative employer, and • to fix his own hours of 'work' within certain parameters and to tailor his time to suit his personal convenience rather than solely the interests of the putative employer, provided that, first, the independence of that person does not appear to be fictitious and, second, it is not possible to establish the existence of a relationship of subordination between that person and his putative employer. However, it is for the referring court, taking account of all the relevant factors relating to that person and to the economic activity he carries on, to classify that person's professional status under Directive 2003/88. 126 Based on this ruling and statements, the self-employment contract of food delivery riders will be regarded as valid, at least in the Working Time Directive context. In other words, similarly to B, riders could continue to be self-employed. Indeed, food delivery workers have a similar amount of latitude as B had. Indeed, the degree of independence, subordination and direction is of greater importance according to the CJEU's guidance on how a worker is defined. The Yodel ruling suggests that the importance of the 'use of substitutes and contractors' is only one part of the puzzle. Indeed, if a subordinate and dependent relationship exists even when substitution is used, a person is likely to fall within the CJEU worker definition. By failing to answer the question on the calculation of the working time, the Court made a conceptual mistake. As Gramano noted, 'The silence of the Court on this matter can be perceived as an implicit, and yet clear, position of the Court, which still relies on an anachronistic view of subordination.' 127 Interestingly, the CJEU follows a similar reasoning that the UK High Court; the possibility of providing subcontractors or substitute proves the self-employed status. In fact, this case is broadly in line with 'sham' test established in the Autoclenz. Therefore, although the UK has left the EU, this ruling will affect the remaining Members, bringing a UK perspective and arguments.
Fortunately for riders, there are at least two aspects that would save them. First, the ruling is deeply influenced by the underlying facts and Yodel's organizational model. However, Yodel's business model is fully in line with many models in the platform economy. 128 Second, the sentence 'the independence of that person does not appear to be fictitious' was discussed in previous cases where different criteria were established to determine whether a self-employment contract was genuine. 129 In fact, Yodel slightly departs from that previous case. 130 According to the FNV ruling, a self-employed contractor should be viewed as a worker 'if he does not determine independently his own conduct on the market, but is entirely dependent on his principal because he does not bear any of the financial or commercial risks arising out of the latter's activity and operates as an auxiliary within the principal's undertaking'. 131 FNV could, however, be regarded as an exception. Indeed, the CJEU did not refer to the 'typical' tests of control and subordination but instead based its ruling on the notion of independence on the market. Interestingly, looking at the CJEU decision in Uber, one could argue that the drivers are an auxiliary within the principal's undertaking. 132 However, it seems less likely that delivery workers will meet this threshold, especially after Yodel.
The need to look beyond strict criteria is confirmed by the Court when it stated: the status of 'worker' within the meaning of EU law is not affected by the fact that a person has been hired as a self-employed person under national law, for tax, administrative or organizational reasons, as long as that person acts under the direction of his employer as regards, in particular, his freedom to choose the time, place and content of his work, does not share in the employer's commercial risks, and, for the duration of that relationship, forms an integral part of that employer's undertaking, so forming an economic unit with that undertaking. 133 In most cases, the Court also found that another essential feature is 'a certain period of time one person performs services for and under the direction of another person in return for which he receives remuneration'. 134 This creates a 'hierarchical relationship between the worker and his employer'. 135 This element of 'direction' may also refer to aspects of control and subordination as established in Dita Danosa v. LKB Lī zings SIA. 136 Indeed, a person performing services 'for and under the direction of another person' in return for remuneration constitutes the 'essential feature of an employment relationship'. 137 These two elements are at the centre of the debate at national level.
Finally, a person is a worker within the meaning of EU law 'as long as that persons acts under the direction of his employer as regards, in particular, his freedom to choose the time, place and content of his work, does not share in the employer's commercial risks, and, for the duration of that relationship, forms an integral part of that employer's undertaking, so forming an economic unit with that undertaking'. 138 While food delivery riders can, to some extent, determine their own schedule, their working hours depends on the restaurants opening hours; it is less evident whether they share the employer's commercial risks. Moreover, they are part of the delivery companies' undertaking. Indeed, Deliveroo and Uber Eats are unable to provide their services without relying on external riders. Those riders are, therefore, part of the companies undertaking. At the same time, the riders can deliver an order for Uber Eats wearing Deliveroo uniforms because of the ease to switch from a platform to the other. However, some of the elements discussed in this case were not taken into consideration in the 2020 judgment, and it can be wondered whether the Court will follow the latest judgment or the long list of precedents.

The way forward?
The Yodel judgment contrasts with recent decisions at national level and raises questions regarding the outcome of future rulings but also the application of specific laws such as the Riders law. Indeed, applying Yodel's criteria to delivery riders, platform workers seem excluded from the scope of application of the Employment Directives. The Yodel ruling could result in various interpretations by national courts, especially the sentence 'the independence of that person does not appear to be fictitious'. Indeed, it seems unlikely that the Spanish legislator will agree to set the brand-new Riders law aside, although it seems to directly conflict with Yodel, as European Union does not have exclusive competence in labour matters. This case brings to light the potential conflicts between the CJEU worker definition and that adopted at national level, which then puts harmonization in jeopardy.
The Commission is expected to publish recommendations on the status of food delivery rider by the end of 2021. Hopefully, these recommendations will be in line with the some of the existing systems. The lack of EU involvement could result in an hinderance of free movement of workers as well as further abuses. Indeed, it does not seem impossible for food delivery status to contract riders in a country offering less protection. For instance, Portuguese riders could easily service close towns on the Spanish side. Similarly, riders could be hired under foreign contracts, as Ryanair does in the aviation sector for instance.
Finding a solution requires a fine balancing exercise to avoid limit the gig economy and its flexibility unnecessarily. At the same time, riders are entitled to some basic protections. Maybe a way forward would be to create a new category, as the Turin Court of Appeal suggested, which would allow for some protections while maintaining some flexibility. A system like the SMart in Belgium also offers great advantageous; it is an opt-in system. Therefore, riders who want flexibility could still remain self-employed. In contrast, those who want a level of protection could become 'employees'. Similarly, recognizing a third category of workers, who are afforded basic labour rights, is a good option. Those middle-ground approaches might be more efficient than recognizing riders as workers. The major disadvantage of granting a worker status to riders is the hindering of flexibility; some riders need flexible hours/days for personal reasons. There have been reports of chronically ill riders who favoured the current system as it allowed them to work when they felt able to but did not require them to call in sick.
The sentence that best summarizes the need for a new approach was given by a US District Judge in the case of Lyft. He noted that 'The test the California courts have developed over the 20th Century for classifying workers isn't very helpful in addressing this 21st Century problem. Some factors point in one direction, some point in the other, and some are ambiguous. Perhaps Lyft drivers who work more than a certain number of hours should be employees while the others should be independent contractors. Or perhaps Lyft drivers should be considered a new category of worker altogether, requiring a different set of protections.' 139 Although it is a US judgment, interestingly, some national courts reached the same conclusions; food delivery riders belong to an in-between category. 140 The EU could follow these examples while regulating the status of riders. This could be the best way forward by granting some basic workers' rights while retaining some flexibility for the platforms: the creation of a hybrid category of self-employed. main issues would be if there is a great difference in protection afforded to employees and this in-between approach, companies could be tempted to make this in-between category the standard one. 141

Conclusion
The gig economy has challenged existing laws at both national and European level requiring new classifications and definitions of the terms 'self-employed' and 'workers'. Indeed, the platform economy offers more flexible work organization while eroding the formal rules in place. The increasing spread of non-standard employment and bogus self-employment status has prompted the debate on how to reshape labour laws to accommodate these new formats.
While the employment status of Uber's drivers is nearly an open-and-shut question around Europe, for food delivery riders, the question is much more complex. Indeed, the significant difficulty with delivery riders is the possible account renting. The possibility of subcontracting is much more straightforward and easier to establish, resulting in some courts being reluctant to grant a worker status. Although Yodel represented a new opportunity for reasoning on the notion of 'worker', the ruling raises more questions than answers. This is a missed chance for the CJEU to become a key player in the debate on the alleged ineffectiveness of the existing categories. After this ruling, it can be expected to encounter even more cases due to the 'high risk of misclassification by means of carefully tailored contractual clauses often aimed at circumventing the application of employment laws, or, the insufficient scope of employment laws to cover new forms of work at the periphery of the notion of subordination'. 142 The decisions in the UK and by the CJEU in Yodel clearly highlight the fine line between employees and self-employed. This might explain the existing divergences and the fact that some countries show an unwillingness to investigate the employment status question. Regulating the work of all delivery people, not just bike riders, requires, therefore, a very fine balance.
Despite the difficulties, regulation is urgently needed. The lack of regulation coupled with the imbalance created by the platforms' structure has resulted in an increasing number of case law being brought at national level. Therefore, national courts in many civil law countries have become the last gatekeeper to avoid abuses on the platforms' side.
The slowness of the response by legislators in Europe, both at national and EU levels, has allowed gross exploitation of labour forces and increased employment precarity. Through their unwillingness to find a solution, States are indirectly endangering the riders' lives. Indeed, workers are encouraged to reach targets which can push them to commit some delicts. 143 While some might argue that committing a delict is a personal decision, such a statement would not consider the pressure those riders are facing. It could also be assumed that if those riders had a fixed salary or at least a decent wage, they would not expose themselves to such risks. It would also help fight social security evasion. 144 Consequently, finding a solution is urgent, especially after the COVID pandemic has demonstrated their importance.
Finding a solution requires a fine balancing exercise to avoid limit the gig economy and its flexibility unnecessarily. At the same time, riders are entitled to some basic protections. Maybe a way forward would be to create a new category, as the Turin Court of Appeal suggested, which would allow for some protections while maintaining some flexibility. A system like the SMart in Belgium also offers great advantages; it is an opt-in system. Therefore, riders who want flexibility could still remain self-employed. In contrast, those who want a level of protection could become 'employees'. Similarly, recognizing a third category of workers, who are afforded basic labour rights, is a good option. Those middle-ground approaches might be more efficient than recognizing riders as workers. The major disadvantage of granting a worker status to riders is the hindering of flexibility; some riders need flexible hours/days for personal reasons. There have been reports of chronically ill riders who favoured the current system as it allowed them to work when they felt able to but did not require them to call in sick.