Abstract
The present article offers an analysis of some key aspects of the UK Supreme Court (SC) Deliveroo judgment. After a short description of some of the facts and findings of the case, the article argues that the Supreme Court may have actually misconstrued the personal scope of application of Article 11 ECHR and, like the other domestic jurisdictions before, misapplied the law (and the concept of ‘employment relationship’ deployed by the ECtHR) to the facts of this case. While the SC judgment did not expressly elaborate on the domestic ‘worker’ definition contained in s. 296 TULRCA 1992, the article explores the extent to which the Deliveroo saga has incorrectly construed this concept, embracing a very narrow concept of ‘personal work’ that neither the statutory wording itself nor the context in which it was applied arguably support. Finally, the concluding section of this article offers an alternative approach to the legal construction and legal regulation of ‘personal work’, one that is already emerging in other jurisdictions and that should underpin any future reform of the personal scope of application of UK, but also EU labour law - a reform, the article concludes, that is long overdue.
Keywords
Introduction
On 21 November 2023, the UK Supreme Court (SC) delivered its long-awaited decision in the Deliveroo case. 1 The judgment concluded that the Deliveroo riders organised and represented by trade union IWGB were not in an ‘employment relationship’ for the purposes of Article 11 of the ECHR. Fatal to this claim was, in particular, the finding that ‘[t]he power conferred on Riders under the new contract to appoint a substitute is virtually unfettered’ (paragraph 69 of the judgment) – in other words, that their power to arrange for a substitute rider to make a delivery on their behalf was incompatible with their being ‘workers’ of the platform. Paragraphs 71 and 72 of the decision reinforced this finding by reference to a number of other factors and indicators emerging from the contractual relationship between Deliveroo and its riders, that equally denied the possibility of them being in an employment relationship since the Supreme Court understood them as ‘free to reject offers of work, to make themselves unavailable and to undertake work for competitors’ (paragraph 72).
This decision puts an end to a domestic judicial ‘saga’ that had started in November 2017, when the Central Arbitration Committee 2 (CAC) refused to issue a declaration of recognition in favour of IWGB under Schedule A1 to the Trade Union and Labour Relations (Consolidation) Act 1992 (TULRCA, often referred to as the ‘statutory recognition procedure’), which would have rendered the union statutorily entitled to negotiate with the employer on pay, hours and holidays, on behalf of the riders working in the London area of Camden. The CAC had refused on the basis that the riders did not fall under the ‘limb (b)’ worker definition set out in s.296 (1)(b) TULRCA 1992 (an individual who works under a ‘contract whereby he undertakes to do or perform personally any work or services for another party to the contract who is not a professional client of his’) because ‘In light of [the] central finding on substitution, it cannot be said that the Riders undertake to do personally any work or services for another party’ (paragraph 101 of the CAC decision). In the same decision, the CAC also rejected the alternative argument that a refusal to recognise IWGB for collective bargaining based on the definition of ‘worker’ in the domestic legislation would constitute a breach of Article 11 ECHR, affirming that ‘on the specific facts of this case and the unfettered and genuine right of substitution that operates both in the written contract and in practice, the argument does not succeed’ (paragraph 104 of the CAC decision). It was only the Article 11 argument that was allowed to proceed through the ensuing judicial review and appellate process, and this explains why the Supreme Court decision is primarily concerned with ECHR rather than domestic provisions and case law. The CAC decision, its judicial review by the High Court, and the subsequent Court of Appeal judgment have already received considerable scholarly attention. 3
The present article offers an analysis of some key aspects of the SC judgment. After a short description of some of the facts and findings of the case in sections 2, 3, 4 and 5, section 6, it argues that the Supreme Court may have actually misconstrued the personal scope of application of Article 11 ECHR and, like the other domestic jurisdictions before, misapplied the law (and the concept of ‘employment relationship’ deployed by the ECtHR) to the facts of this case. While the SC judgment did not expressly elaborate on the domestic ‘worker’ definition contained in s. 296 TULRCA 1992, section 7 of this article will explore the extent to which the Deliveroo saga has incorrectly construed this concept, embracing a very narrow concept of ‘personal work’ that neither the statutory wording itself nor the context in which it was applied arguably support. Finally, the concluding section of this article will offer an alternative approach to the legal construction and legal regulation of ‘personal work’, one that is already emerging in other jurisdictions and that should underpin any future reform of the personal scope of application of UK, but also EU, labour law 4 - a reform, the article concludes, that is long overdue.
Some aspects of the judgment, for example those pertaining to the compulsory nature of the bargaining process envisaged by Schedule 1 of TULRCA 1992 or whether Article 11 ECHR should be understood as conferring on workers a right to require the employer to bargain collectively with their union, are not considered.
The key facts of IWGB v Deliveroo
IWGB first started organising Deliveroo couriers early in 2016. In November 2016, after an unsuccessful campaign aimed at agreeing recognition on a voluntary basis with Deliveroo, IWGB sought to trigger the statutory recognition procedure provided by Part I of Schedule A1 TULRCA 1992. The procedure, if successful, can lead to the CAC issuing a declaration of recognition, effectively entitling the recognised union to negotiate with the employer on pay, hours and holidays. For this to happen, the applicant union needs to navigate successfully a number of substantive and procedural hurdles 5 and, most importantly, either succeed in establishing that it represents more than half of the workers in the bargaining unit or, if there are doubts about this level of representativity, to win a secret ballot held by the CAC and confirming support by a double majority of the workers participating in the ballot and at least 40% of the workers constituting the bargaining unit. Throughout this lengthy and cumbersome process, the meaning of the word ‘workers’ is the one specified by s.296 (1)(b) TULRCA 1992 a wording similar but, as shall be pointed out in section 7 below, not identical to the better known s.230(3)(b) ERA 1996 ‘limb (b)’ worker definition. 6
In the course of the entire process, one of the objections raised by Deliveroo was that IWGB did not and could not represent any ‘workers’ in the unit as none of the 10,808 riders the company engaged to perform deliveries to customers, whether by motorbike or bicycle, were workers within the meaning of s.296 of the Act, being instead ‘suppliers’. After various procedural requirements were dealt with, the application by IWGB went to a full hearing before the CAC at the end of May 2017.
At the hearing it became clear that the contractual terms applying to the couriers were, by and large, similar to those prevailing in the gig economy and in the urban transport and food delivery sector. They will not be discussed in this article and can be found, in summary, at paragraphs 21–30 of the Supreme Court judgment. There was, however, a contractual term that stood out of the more mundane array of clauses used by platform employers: Clause 8. A very broad – and recently introduced - ‘substitution clause’, that ended up playing a crucial role in supporting the platform's argument that the couriers were not workers but, instead, suppliers.
Clause 8: The new, broad, substitution clause
It is worth stressing that between IWGB starting to organise Deliveroo couriers in 2016 and the CAC hearing in May 2017, two important developments had taken place. Firstly, in October 2016, the London Employment Tribunal released a judgment in Aslam v Uber finding that the 30,000 drivers working for Uber were indeed its ‘workers’ according to the definition of the term contained in s. 230(3)(b) ERA 1996, 7 a definition that is very similar but, crucially, not identical to the one contained in s.296 (1)(b) TULRCA 1992 . Secondly, ‘Shortly before the first hearing before the CAC, Deliveroo introduced new contracts’ (on 11 May 2017 – paragraph 23) and these new contracts contained a new clause, now (in)famously known as ‘Clause 8’.
The new Clause 8 set out the provisions concerning the right to appoint a substitute in very broad terms, stating that: ‘Deliveroo recognises that … you may wish to engage others to provide the Services. Deliveroo is not prescriptive about this and you therefore have the right, without the need to obtain Deliveroo's prior approval, to arrange for another courier to provide the Services (in whole or in part) on your behalf. This can include provision of the Services by others who are employed or engaged directly by you’ (Clause 8.1).
It is not academic to mention this particular succession of events, nor the exact wording of Clause 8 - which is likely to become far more popular in the future as a ‘defeat device’ designed to deny employment status to UK workers in the gig economy, and possibly beyond. It is also worth recalling that in February 2017, three months before the Deliveroo hearing and just weeks before Deliveroo introduced Clause 8, a ‘lesser’ substitution clause contained in the Pimlico Plumbers contracts 8 had been found by the Court of Appeal in Pimlico Plumbers v Smith 9 to be too constraining, and not unfettered enough, to deny ‘worker’ status. 10
Only a cynic would think that Clause 8 was introduced for the purpose of avoiding a finding similar to the one in Uber, and that its wording was deliberately loosened with a view to expanding the terms of the substitution clause contained in the Pimlico contracts.
The effects of the substitution clause
Clause 8 achieved successfully the purpose it had been designed for (but then, as the CAC pointed out, ‘Even if they did it in order to defeat this claim and in order to prevent the Riders from being classified as workers, then that too was permissible’). 11 It did so in all of the various courts that concerned themselves with this dispute, from CAC up to the Supreme Court.
In 2017, the CAC concluded that the Deliveroo riders were not workers under s. 296 TULRCA 1992, to a large extent because: ‘The central and insuperable difficulty for the Union is that we find that the substitution right to be genuine’; 12 and ‘In light of our central finding on substitution, it cannot be said that the Riders undertake to do personally any work or services for another party. It is fatal to the Union's claim’. 13 This ‘genuine’ substitution finding was both deployed to defeat the ‘personal work’ requirement attached to the s. 296 TULRCA 1992 worker definition and, no less importantly, to reject the alternative argument that a refusal to recognise IWGB for collective bargaining based on the definition of ‘worker’ in the domestic legislation would constitute a breach of Article 11 ECHR. On this point CAC affirmed that ‘on the specific facts of this case and the unfettered and genuine right of substitution that operates both in the written contract and in practice, the argument does not succeed’ 14 - a rather unceremonious way to dismiss a serious argument.
Fast forward to 2023 and the Supreme Court essentially confirmed the deliberations of the CAC (as did all previous courts). It is worth recalling that from the CAC hearing onwards, the only legal issue to be tried was the one pertaining to the compatibility of s. 296 TULRCA 1992 (as interpreted by the CAC) with Article 11 ECHR since, as pointed out by the Supreme Court, as soon as IWGB sought to judicially review the CAC decision before the High Court, the judge at the preliminary hearing established that none of the judicial review grounds pertaining to the s. 296 TULRCA 1992 worker definition raised any arguable error of law and, in conclusion, ‘She gave permission for judicial review only on the article 11 ground’. 15 The precise way in which the Supreme Court dealt with the Article 11 and ‘employment relationship’ arguments is discussed at greater length in section 6 below, but essentially it was rejected on the ground that both on paper and (as we shall see in the next section) in practice, the presence and use of the substitution clause was incompatible with any such argument.
Substitution in practice
It is important to acknowledge, from the outset, that there was some material evidence that the substitution clause in Clause 8 had actually been actioned by at least some couriers. Paragraphs 28 and 29 of the Supreme Court judgment identify two such instances.
Firstly, and this is arguably the most challenging factual finding, there was a courier, Asim Munir, that gave evidence before the CAC to the effect that ‘he regularly engaged a substitute, taking 15–20% of the fee he received from Deliveroo and passing on the balance to his substitute’. 16 This would appear to both contradict a claim of personal performance of work and display a type of entrepreneurial activity, whereby one is organising the work of others or, as the CAC decision noted, ‘exercising the substitution provisions for his own potential profit’. 17 We do not know how regular this practice was nor whether, alongside this activity, he also provided his work personally and with what frequency.
Secondly, there was ‘one instance of substitution after the job had been accepted and before collection of the Order from the restaurant’. 18 There is little analysis of the implication of this finding, but it would have probably been understood as suggesting that even after a particular delivery had been agreed between a rider and Deliveroo, the former could still change his mind and delegate the task to a substitute, the substitution clause overriding the prior agreement between the first parties. It is worth noting that the evidence provided by this courier was challenged by the counsel for the union, Lord Hendy KC, who ‘questioned the plausibility of the account’. 19 The CAC admitted that ‘[i]t does sound a little surprising, but even if the whole situation was crafted to provide an example of a mid-job substitution, it effectively demonstrates the capacity of a Rider to do such a thing, should they want to’. 20
Thirdly, there was a survey conducted by Deliveroo that was circulated during the CAC hearing suggesting that ‘14 of the 65 Riders who answered the question had either themselves used a substitute or knew of other Riders who did’. 21 That appears to have attracted no interest before the Supreme Court - unsurprisingly one would say, given the obvious methodological flaws in terms of sampling and imprecise questions, based more on hearsay than facts. 22
It is important to put these numbers into context. Two couriers out of 100 operating in the proposed Camden bargaining unit were making some use of Clause 8.
23
That is by all accounts a small, negligible number that, on its own, should have suggested caution in making a more generalised inference about the employment status of all other workers. As confirmed by the Supreme Court ‘[t]he CAC found that “a few, if that, Riders use substitutes”. Most Riders did not use a substitute as they did not need to do so.’
24
. Nevertheless the SC proceeded to conclude that ‘Such a broad power of substitution is, on its face, totally inconsistent with the existence of an obligation to provide personal service which is essential to the existence of an employment relationship within article 11. (See ILO Recommendation No 198, as adopted into the Strasbourg jurisprudence by The Good Shepherd.)’
25
The Article 11 ECHR ‘employment relationship’ question reassessed: a case of ‘contractual bias’ in the application of ‘relational principles’?
The Supreme Court chose to approach the question of whether the couriers were covered by Article 11 of the Convention by reference to the concept of ‘employment relationship’ arising from ILO Recommendation 198. In doing so, the UK Supreme Court could rely on the 2013 Grand Chamber judgment on the employment status of members of the Romanian clergy for the purposes of establishing a union (Sindicatul “Păstorul Cel Bun” v Romania) 26 that had indeed deployed the concept of employment relationship as emerging from ILO Recommendation 198 to define the scope of Article 11 ECHR.
John Hendy KC, counsel for the union, presented to the SC the alternative option to construe the scope of Article 11 by reference to the broader proposition that ‘the right to bargain collectively [be] enjoyed by every individual with an occupational interest to protect’, 27 finding some authority in the obiter dicta of another ECtHR decision of 2015, 28 where a case involving self-employed farmers whose attempts to register a trade union under Romanian law were frustrated by the national authorities on the ground that self-employed workers could not form unions but only join pre-existing ones. In Manole, the ECtHR noted that ‘Article 11 § 2 of the Convention does not exclude any occupational group from the right of association’ and that ‘insofar as the applicants were refused the right to be registered as a trade union-type association… there was an interference by the respondent State with the exercise of the rights guaranteed by Article 11 of the Convention’. However, it also found that this interference could be justified under Article 11(2) insofar as it could be seen as a proportionate way to pursue the legitimate aim to protect ‘the economic and social order by maintaining the legal difference between trade unions and other types of association’, 29 having also considered various ILO sources and observations and concluded that there were ‘no sufficient reasons to deduce from those general observations that the exclusion of self-employed farmers from the right to form trade unions amounts to a breach of Article 11 of the Convention’. 30 This was more than enough for the UK Supreme Court to dismiss the argument altogether, claiming that ‘Manole does not detract from the clear authority of the Good Shepherd, a decision of a Grand Chamber, to the effect that the trade union right under article 11 arises only in the context of an employment relationship’. 31 This dismissive approach by the SC has already been criticised by Professor Keith Ewing, in a recently published blog post. 32
Even conceding that the Supreme Court, and the previous jurisdictions before it, correctly identified the applicable law and the relevant principles contained therein, they have, it is submitted, applied them incorrectly. Or rather, they have applied them with a – perhaps unconscious – contractual bias that has prevented them from fully identifying and recognising the ‘relational’ – as opposed to purely contractual – dimensions of the provision and performance of work under the Deliveroo contract. 33 It is arguable that these relational dimensions should have been taken into account in order to determine the employment status of the couriers, and establish whether they were, in reality, in an employment ‘relationship’ under Article 11 ECHR.
It is hardly a coincidence that the very first provision of ILO R198 offering some guidance on the determination of an employment relationship focuses on the ‘primacy of facts’ principle. Paragraph 9 of the Recommendation demands that such determination be ‘guided primarily by the facts relating to the performance of work and the remuneration of the worker, notwithstanding how the relationship is characterized in any contrary arrangement, contractual or otherwise, that may have been agreed between the parties’. Arguably, this provision asks national authorities to perform an intellectual exercise that is, in many ways, second nature for most continental European judiciaries, but – admittedly – one that runs against the grain of the common law, freedom of contract based, tradition. 34 This intellectual exercise requires the judicial interpreter to abstract from the terms of the contract and focus primarily on the performance of the work by the worker. It is inherently an exercise that requires the interpreter to examine with painstaking detail the way in which each individual worker performs his work to the benefit and for the profit of the enterprise, with little or no preoccupation for the way contractual documents depict that performance on paper.
In stark contrast, and in spite of important legal developments such as the ones embodied by the Autoclenz 35 and Uber 36 judgments, English courts are apt (in fact adept) at performing a different kind of exercise, consisting in identifying working practices and behaviours that are consistent with and – in a rather circular way – capable of simultaneously giving credence and validating the contractual terms agreed, on paper, by the parties. The two types of enquiry are pretty much antithetical and difficult to reconcile – it is an either-or type of intellectual exercise. There is no shortage of examples in UK labour law of the manifestation of this type of fallacy. The ‘necessity test’ that needs to be satisfied to imply a contractual relationship between a user and a temporary agency worker (something possible only when strictly indispensable to give reality to a relationship that cannot otherwise be explained in the alternative by any of the existing contractual documents, e.g., James v Greenwich) 37 is a prime example of this contractualist mindset.
Even in the Deliveroo case, there are several instances suggesting that the UK judicial interpreters involved therein have been the – perhaps unconscious – victims of this contractual bias, in reality a fallacy. There is arguably no clearer example of this fallacy than the CAC panel's way of dealing with the case of the ‘mid-job substitution’ performed by the rider who had already accepted a job but decided to pass it to another colleague. At paragraph 81 of the decision CAC's decision, the judge posits that ‘even if the whole situation was crafted to provide an example of a mid-job substitution, it effectively demonstrates the capacity of a Rider to do such a thing, should they want to’ (emphasis added). From an English contract law point of view, this may be an eminently correct way to proceed, and a reflection of the aforementioned tendency to look into factual practices to validate and give credence to contractual terms. However, from a relational point of view, this approach leads to false positives and incorrect determinations.
A judge trained in a more relational approach would make little or nothing of the finding that a contractual clause gives the capacity to a subordinate worker to act as a self-employed person or, in the case of the courier that demanded a fee from his substitutes, in a business/employer-like manner. Certainly, a civil law and, arguably, a human rights judge would not draw the conclusion that if one or two couriers have indeed actually actioned a clause that allows them to perform their economic activity in a business-like manner, this also affects the relationship of all the other workers covered by the same contract but whose way of performing their own work has not involved making any use of that contractual clause, which is - as far as they are concerned – a dead letter, entirely eclipsed by their primacy of facts.
This contractual tunnel vision carries through across the various levels in the Deliveroo saga, all the way up to the Supreme Court judgment. For instance, their Lordships contend that the ‘[CAC] found that Deliveroo did not object to the practice of substitution by a Rider for profit or to Riders working simultaneously for competitors of Deliveroo. In all the circumstances, the CAC was entitled to conclude that the contractual provisions genuinely reflected the true relationship’. 38 It is debatable whether such a finding made in a continental European court would cut to the chase in respect of the actual workers who had actually made some use of the substitution clause. In fact, we know that similar substitution clauses were altogether dismissed as irrelevant by Dutch courts. 39 In the Dutch version of the Deliveroo saga, the equivalent to Clause 8 stated that ‘[u]nder the contract, delivery persons could be replaced to have the delivery carried out by someone else, provided that the replacement showed a valid ID and proof of the right to work in the Netherlands before starting work’. 40 The Hoge Raad considered this clause and found that substitution also ‘happens in practice’. 41 However, noting that the technology itself and the fact that the app had to be passed to the substitute made it impossible ‘for a delivery person to … have his work carried out by all kinds of other substitutes at the same time and for this to become a revenue model’, it concluded that ‘[t]he replacement option that delivery drivers have is … not incompatible with the existence of an employment contract, as within an employment contract it is also possible for the employee to be replaced with the employer's permission’. 42
Arguably – in a continental court - even a finding that those riders systematically using substitutes were self-employed, it would not have had any bearing in terms of characterising the relationship of all the other workers that never made any significant use of the substitution clause (even if they had the contractual ‘capacity’ to do so, paraphrasing CAC) as a contract for services or as a business undertaking of sorts. The contractual provisions may well have ‘genuinely reflected the true relationship’ of those two workers that used Clause 8 in practice, but suggesting that this also meant that every other worker, regardless of their true relationship, was equally affected by this contractual classification would be a non-sequitur under most jurisdictions, including arguably under the jurisdiction of the ECtHR, a judicial body whose modus operandi makes it particularly sensitive to assessing claims on an individual basis.
In the 2008 publication Towards a Comparative Theory of the Contractual Construction of Personal Work Relations in Europe, 43 Mark Freedland and I advanced the hypothesis that – in the area of labour contract regulation – it was possible to contrast between English common-law-based systems and continental European civil law-based systems, by reference to the former being characterised by a ‘regulated self-designed contracts’ approach, and the latter by a ‘standardised contract typology’ approach. Both legal culture and legal institutions appeared to play a key role in the emergence of this contrasting approach. The regulated self-designed contracts approach bestows upon its parties a much greater degree of flexibility in terms of the clauses that can be added to it and actioned upon, as well as – related to that – in terms of the classification of the contractual arrangements it is capable to generate.
The meeting of the ‘regulated self-designed contract’ tradition with work provided and supervised through digital platforms has further exacerbated the self-designed features of the English approach. By allocating a number of managerial functions to the algorithm, including managing the way in which work is offered, accepted, performed, supervised and evaluated, employers have been increasingly able to draft contracts as seemingly neutral governance mechanisms that offer a selection of seemingly infinitely flexible regulatory options for the parties to pick and choose from. Any business risk potentially deriving from a worker opting out of the various contractual features characterising employment (control/subordination, integration, risk of loss) is more than mitigated by the constant monitoring, evaluation and even disciplinary capacity of the algorithm.
It would no doubt be an oversimplification to redefine these two approaches as, respectively, an ‘à la carte’ approach for the UK and a ‘set menu’ for the civil law traditions, but in the Deliveroo decisions there is almost a sense that English judges have indulged in assuming that because a handful of diners had ordered the vegetarian options from a uniquely vast ‘à la carte’ menu then surely, that menu having the capacity to be a vegetarian one, every other diner in that establishment was also a vegetarian. Contractualist biases aside, there is a non-sequitur here. This approach is clearly not delivering in terms of the relational approach required to assess the compatibility of UK law with Article 11 of the ECHR.
A better approach, and one that the Strasbourg court will undoubtedly be asked to consider carefully, would have been to conclude that if the couriers that had actually used the substitution clause were to fall outside the ‘employment relationship’ definition of R198, then this should have been of no consequence for the classification of all the other couriers whose actual work performance pointed, in practice, towards an employment relationship as defined by R 198 and protected under Article 11. The easier solution, in domestic proceedings, would have been to expunge those two couriers from the bargaining unit calculus. However, it is worth remembering that other domestic courts, such as the Dutch Supreme Court in its own Deliveroo judgment of March 2023, have gone beyond that and have simply ignored the impact of substitution clauses even when their use was relatively widespread. 44
Personality at work reassessed 45
Regardless of the fate any possible reference to the Strasbourg court, it is important to stress that the Deliveroo saga has, in the meantime, cast a long shadow on the concept of personality in work, with significant implications in terms of narrowing the statutory definitions of ‘worker’ deployed in a number of UK Acts of Parliament, first and foremost that of TULRCA 1992. There is a clear sense that contractual substitution terms similar to Clause 8 could, in the coming weeks and months, be retrofitted to large swathes of contractual arrangements for the provision of (platform and non-platform) work, certainly in those where the performance by a specified individual is not material to the enjoyment of the service by a client or customer, and where the algorithmic management features embedded in the app itself are able to monitor, manage and, where necessary, discipline the performance irrespective of the actual person performing it. There will no doubt be a pressure on courts to test the impact of such substitution clauses also on s. 230 ERA 1996 ‘workers’ rights.
This section suggests that this outcome is neither desirable nor, strictly speaking, justified, as – arguably – the concept of personality bestowed by domestic courts to the worker definition contained in s. 296 TULRCA 1992, and to other worker definitions, is unduly narrow.
A first point to clarify is that there is more than one ‘worker’ definition and more than one ‘personal’ work concept in UK labour law statutes. English courts have tended to approach the concept of ‘personality’ contained in the various statutory worker definitions as somewhat of a monolith, whereas it is not. Using the ‘personal work’ requirement as a yardstick, it is arguably possible to identify a spectrum of defining concepts in various UK statutes, ranging from the narrower ‘employee’ concept contained in s. 230(1)-(2) ERA 1996 at the one end, to the much broader concept of ‘home worker’ contained in s. 35(1) of the National Minimum Wage Act 1998 (NMWA) at the other, opposite, end of the spectrum.
Under s. 230(3) ERA 1996, the ‘employee’ concept is linked to the ‘contract of service’ notion, which courts have consistently defined as requiring personal performance, so much so that, according to Peter Gibson LJ, a ‘clause … entitling [a person] not to perform any services personally, is a provision wholly inconsistent with the contract of service’.
46
At the opposite end of the spectrum, the statutory definition of ‘home worker’ does not impose a ‘personal work or service’ requirement at all, with s. 35(1) of the NMWA1998 expressly providing that, for the purposes of that Act, the standard ‘worker’ definition contained in s. 54 shall ‘have effect as if for the word “personally” there were substituted “(whether personally or otherwise)”’. In his judgment in James v Redcats (Brands) Ltd,
47
Elias J persuasively opined that ‘unlike the worker, it is not necessary that the home worker should personally carry out the work at all. It is the provision of services that is enough, irrespective of who provides them, as long as they are not provided in the course of running a business or profession’. This is because, per Elias J, ‘frequently the work will involve, for example, making shoes or clothes, and that work is typically carried out by members of a family or a group of friends. (The growth of computer technology has also spawned many workers operating from home (so called “teleworkers”) although they are more likely to do the work personally.’
48
If the ‘employee’ and the ‘home worker’ definitions represent the two extremes of the ‘personality in work’ spectrum in UK labour law, then it is possible to argue that the two ‘worker’ definitions contained, respectively, in section 230(3)(b) ERA 1996 and section 296(1)(b) TULRCA 1992, sit somewhere within that spectrum. It is also arguable that the construction and literal interpretation of the ERA's ‘limb (b) worker’ definition should point to a ‘worker’ notion that is broader than the ‘employee’ notion, and narrower than the ‘home worker’ notion. A limb (b) worker is defined by the ERA 1996 as someone who works under ‘any other contract …whereby the individual undertakes to do or perform personally any work or services for another party to the contract whose status is not by virtue of the contract that of a client or customer of any profession or business undertaking carried on by the individual’. The Supreme Court judgment in Pimlico Plumbers has confirmed that a clause or practice allowing a certain qualified degree of substitution with workers already employed by the company (the ‘only right of substitution was of another Pimlico operative’) 49 is compatible with the ERA 1996 notion of ‘worker’ as long as it does not provide an ‘unfettered right to substitute at will’. 50
It is also arguable that the TULRCA's ‘worker’ definition has been worded by Parliament as providing a broader, more generous personal scope for the purposes of that Act. The TULRCA worker definition covers ‘any other contract whereby [the person] undertakes to do or perform personally any work or services for another party to the contract who is not a professional client of his’, which – in contrast with the ERA limb (b) definition – does not refer to the ‘business-customer’ exclusion but only to the ‘professional-client’ exclusion. The exact differentiation of these two concepts aside, this is a small but significant difference that incontrovertibly points toward a broader definition of the TULRCA's ‘worker’ concept. With the judgment in Deliveroo, English courts have effectively interpreted the personality requirement in the TULRCA ‘worker’ definition as being identical to the one in the ERA 1996 notion, concluding that ‘the Riders were not workers within the statutory definition of either section 296 TULRCA or section 230(3)(b) of the Employment Rights Act 1996’. 51
Arguably, the use of two ‘worker’ concepts by reference to different definitions denotes the Parliament's intention that these two concepts cover two different work situations including, in my view, different degrees of personality. While it would be inconsistent with the TULRCA's ‘worker’ definition to water down the ‘personal work’ requirement to the point of collapsing that definition into the NMWA 1998 ‘home worker’ definition, it would be appropriate to identify an alternative understanding of ‘personality’ so as to distinguish it from the ‘limb (b)’, ERA 1996, Pimlico Plumber, definition (and of course from the ‘contract of service’ definition).
The ‘personality in work’ requirement of the TULRCA 1992 worker definition should arguably be looser than the Pimlico ‘qualified substitution’ power criterion and, at the same time, tighter than the ‘unnecessary’ Redcats ‘homeworker’ concept. I suggest that any type of qualification that would bring within the concept a personality requirement, however tenuously defined but sufficiently present so as to not make it fall in the ‘unnecessary’ Redcats category, should be sufficient for the purposes of the TULRCA worker definition. 52
From this perspective, it could be argued that the provisos contained in Clause 8.1 of the Deliveroo contract (‘it may not include an individual who has previously had their Supplier Agreement terminated by Deliveroo for a serious or material breach of contract or who …. has engaged in conduct which would have provided grounds for such termination had they been a direct party to a Supplier Agreement’ but one who must have ‘the requisite skills and training’, the right to reside and work in the UK with the right papers, no unspent criminal convictions, and one who will comply with all legal requirements including the Highway Code, all conditions for which the rider was wholly responsible) introduce a sufficient degree of ‘attenuated personality’ as to distinguish this type of arrangement from ‘home work’ arrangements, without, however, qualifying it to the extent needed to meet the stricter threshold of personality expected by Pimlico for ERA 1996 limb (b) workers (which allow for substitutes to be chosen from within the workforce of the company).
There is finally the personal work concept in s.83(2)(a) of the Equality Act 2010, used to define the concept of ‘employment’ for the purpose of equality rights, that ought to apply to any ‘employment under a contract of employment, a contract of apprenticeship or a contract personally to do work’ (emphasis added). This definition visibly omits any of the ‘exceptions’ applying to the ERA 1996 and TULRCA 1992 ‘worker’ concepts and is therefore to be understood as underpinning an even broader concept. Surely, decisions such as Jivraj v Hashwani 53 have cast this concept in the narrowest of terms, excluding the self-employed. But there is now strong authority emerging from the CJEU to the effect that the scope of application of EU equal treatment legislation ought to be understood as applying to ‘a wide range of occupational activities, including those carried out by self-employed workers in order to earn their livelihood’ (emphasis added) and only excluding ‘those consisting of the mere provision of goods or services to one or more recipients and which do not fall within that scope’, 54 casting serious doubts as to the correctness of the UK House of Lord's determination in Jivraj. If self-employed workers are indeed covered by equal treatment legislation, then we might expect an even looser ‘personality at work’ requirement applying to the defining concept framing the personal scope of equality law.
In other words, it is arguable that UK labour law distinguishes (or ought to be understood as distinguishing) between five ‘personality in work’ requirements - a (seemingly) ‘strong personality’ requirement for the ‘contract of service’ category (but even on this see the concluding section of this article); a ‘qualified personality’ requirement for the ERA 1996 ‘limb (b) worker’ category; an ‘attenuated personality’ requirement for the TULRCA 1992 worker category; a ‘loose personality’ requirement for equal treatment legislation; and a ‘no personality’ requirement for the NMWA 1998 ‘home worker’ category, where the power to substitute is unfettered and the personal nature of work is not a necessary requirement - and that the Deliveroo saga judgments have unduly blurred and collapsed the two central ‘worker’ concepts into one.
This suggested approach would also be consistent with the consideration that the TULRCA 1992 worker definition in the context of the application of the recognition procedure contained in Schedule A1 of the 1992 Act is functionally connected to the enjoyment of a fundamental right, the right to bargain collectively, as protected by Article 11 of the ECHR, a number of fundamental ILO Conventions, and Article 6 of the European Social Charter. In December 2018, a few weeks after the High Court judgment in IWGB, the European Committee of Social Rights published its decision in Irish Congress of Trade Unions (ICTU) v. Ireland (Complaint No.123/2016), declaring that a prior exclusion of Irish self-employed workers from the right to bargain collectively was in breach of the European Social Charter.
This decision put in motion a series of events 55 that eventually led to the adoption, at the EU end of the process, of the Communication from the Commission Guidelines on the application of Union competition law to collective agreements regarding the working conditions of solo self-employed persons (2022/C 374/02), exempting some ‘solo self-employed’ covered by collective agreements from the reach of EU competition law. In these guidelines – a clear attempt to reverse the restrictive impact of the CJEU judgment in Case C-413/13, FNV Kunsten Informatie en Media - a ‘solo self-employed’ is defined by the Guidelines as ‘ a person who does not have an employment contract or who is not in an employment relationship, and who relies primarily on his or her own personal labour for the provision of the services concerned’ (emphasis added), the term ‘primarily’ being chosen (consciously, I would suggest) instead of the term ‘exclusively’ in order to encompass both some use of capital in the provision of work and some degree of (occasional) substitution in the provision of one's labour, a practice not uncommon in the various forms of predominantly personal work relations to which the Guidelines apply.
All along this process, the ILO and a number of its interpretative and supervisory bodies have reinforced the view that self-employed workers are entitled both to freedom of association and to the right to bargain collectively. 56 With the CEACR stating – most recently – that Convention 98 ‘only provides for exceptions to its personal scope of application in respect of the armed forces and the police (Article 5) and public servants engaged in the administration of the State (Article 6), and that it therefore applies to all other workers, including self-employed workers’ 57 and inviting Member States ‘to ensure that all workers covered by the Convention, irrespective of their contractual status, are authorized to participate in free and voluntary collective bargaining … [and] … identify the appropriate adjustments to be made to the collective bargaining mechanisms in order to facilitate their application to the various categories of self-employed workers’, 58 there are very strong arguments for the fundamental rights of freedom of association and collective bargaining to apply beyond employment contracts and the employment relationship and to cover all those performing predominantly personal work.
Conclusions: The legal regulation of personal work relations
The word ‘personal’ is materially present in the legal definitions shaping the personal scope of application of a number of UK labour law statues. 59 It cannot simply be ignored by judicial interpreters and advocates or swept under the carpet by academic commentators as unlike, say, concepts such as ‘mutuality of obligation’, it is not the result of a judicial gloss being applied to statute. It is statute.
It also serves a purpose. Work entirely devoid of any personality element risks, in the extreme, rendering the category of work itself over-inclusive if not meaningless, for instance by including those whose work is the manifestation of a power over the use of other people's work, i.e., those who hire staff in the course of running a business or profession – in other words, ‘employers’. It is therefore a safeguard against the collapse of the category of labour into that of capital. In that sense, one could remain attached to, and continue to justify, the ideas of ‘personality in work’ posited in previous co-authored work, such as the 2011 monograph with Mark Freedland ‘The Legal Construction of Work Relations’ (where it was defined as ‘mainly’ (emphasis added) personal) or the 2019 ETUC paper co-authored with Valerio De Stefano ‘New trade union strategies for new forms of employment’ (describing it as ‘predominantly’ (emphasis added) personal). In the meantime, as noted above, the idea of personal work relations (and akin or cognate concepts) has been deployed in a number of guises in order to expand the personal scope of application of various areas of labour 60 and equality law, 61 both validating the idea of the legal construction of personal work relations and increasingly shaping an argument for their legal regulation.
However, personality at work should not be fetishised either, and it should certainly not become the new legal technique for replacing the stranglehold so far exercised, in certain legal systems, by ‘mutuality of obligation’ on the concept of employment, especially now that the days of ‘mutuality’ seem to be numbered. 62
The very week the UK Supreme Court published the judgment in Deliveroo, several academics (certainly those at University College London) would have received an email from their line managers warning them that the flu season had started, and that if lecturers felt unwell and could not deliver a class, before rescheduling (an increasingly onerous task in our overcrowded universities) they should ‘reach out to the teaching team … to see whether there is any possibility of cover for the lecture/seminar/tutorial within the module teaching team’, an express invite to explore a substitution arrangement. In spite of the very personal nature of academic work, and lecturing in particular, it has long been customary to find substitutes (from within the teaching team, but also outside of it – no line manager has ever objected to, say, a QC or KC delivering a guest lecture) to provide cover in case of illness, sudden caring obligations, but also when important ‘policy impact’ opportunities or conference opportunities materialise. Academics, of course, are also accustomed to working and teaching for more than one employer (in the absence of exclusivity clauses in their contracts), to applying and receiving grants that (with the encouragement of their academic institutions) they use to set up research projects and centres and, indirectly, employ people, and – from setting spin-offs to practicing a liberal profession - are no strangers to all sorts of entrepreneurial activities. Never have these practices been exploited to challenge their status as employees, perhaps suggesting that even that notion does not require in practice – and does not necessitate in law - a strict ‘personality’ element. Similar attitudes are of course prevalent in many other professions, and are likely to become more common as we move increasingly further away from industrial labour and towards an industrious society, where some workers make the most of the technological opportunities presented to them in terms of remote work and juggling multiple jobs and occupations at the same time. Obviously, substitution clauses and the absence of exclusivity requirements appear to be far less straightforward – in the eyes of employers and the UK judiciary at least - when they are used by working class people.
Technological developments have shaken the kaleidoscope of work relations. In particular, the possibility of providing labour through platforms or remotely through ‘telework’ arrangements (as presciently defined by Elias J in Redcats) has transformed several workers into the modern equivalent of the traditional concept of ‘home workers’ who can perform their work away from the employer's premises, but – unlike ‘home workers’ – while very much remaining under the employer’s digital supervision and algorithmic control. As insightfully noted a few years ago by Aurélien Acquier: ‘the emergence of platform capitalism should be understood as a digital reincarnation of the “putting-out system” a pre-industrial organizational form that preceded the emergence of manufacturing and the managerial corporation. In this system, merchants outsourced work to individuals who produced goods at home and owned their own means of production.’
63
Footnotes
Declaration of conflicting interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
