Employees , workers and the ‘ sharing economy ’ Changing practices and changing concepts in The United Kingdom

Recent years have seen a radical shift in the practice and profile of the labour economy in the United Kingdom consisting in the considerable growth of the so-called ‘Sharing Economy’ or ‘Gig Economy’, better identified as the ‘On-demand Economy’. From that starting point, it is argued that a corresponding change seems to have occurred in the set of concepts which the labour/ employment law of the United Kingdom uses to analyse and to characterize the work relations and work contracts which are created, made, and operated within this rapidly growing sector of the labour market. Two recent high-profile Employment Tribunal decisions in the Uber and Citysprint cases, and a decision of the Court of Appeal in this same area in the Pimlico Plumbers case have served to confirm the legislative creation of a third intermediate category of ‘workers’ who benefit from a set of employment rights which is more limited than that enjoyed by employees but which is nevertheless very important. This crystallization of labour law’s newly tripartite taxonomy of work relations has occurred very largely in the context of the on-demand economy, and is beneficial to those located in that sector. This is, however, a rather fragile conceptual structure.

This paper takes as its starting point a radical shift in the practice and profile of the labour economy in the United Kingdom consisting in the considerable growth of the so-called 'Sharing Economy' or 'Gig Economy', better identified as the 'On-demand Economy'.From that starting point, it is argued that a corresponding change seems to have occurred in the set of concepts which the labour/ employment law of the United Kingdom uses to analyse and to characterize the work relations and work contracts which are created, made, and operated within this rapidly growing sector of the labour market.UK employment law has traditionally used the same bi-partite set of categories as those of tax and social security law to classify and to differentiate personal work relations and contracts, namely by distinguishing be-

The Rise of the On-Demand Economy in the United Kingdom
The phenomenon under scrutiny in this contribution is often traced back to the concept of 'crowdsourcing', as described in technology magazine Wired just over ten years ago. 1 Under varying (yet overlapping) labels such as 'crowdwork', the 'gig economy', or 'sharing economy', a new industry of digital platforms has sprung up in the intervening decade, relying on websites and / or mobile phone applications ('apps') to intermediate between customers and large pools of 'on-demand' workers. 2 Originating mostly in the United States, most of the major operators have today found a firm foothold in the United Kingdom.At first glance, we observe a near-unlimited number of variations in business models: some platforms focus on particular services (the most frequently-discussed ones being transportation apps Uber or Lyft, the latter of which has however not yet established operations in the United Kingdom), whilst others deliver their services exclusively online (such as digital task platforms Mechanical Turk or UpWork), whilst yet others deliver a wide range of services -from delivering food to assembling modular furniture -in major urban areas (this includes Deliveroo for food delivery across major towns and cities in the UK, or TaskRabbit for a wide range of domestic services in London).Amidst this variety of operators and business models, there is one constant: providers' reliance on a large and constantly available workforce, to which individual tasks are assigned through sophisticated algorithms driven by consumer demand.
It is for this reason that we eschew the terminology of the 'sharing' or 'gig' economy in this contribution, relying instead on the terminology of an 'on-demand' economy, for two reasons.Despite much industry hype and lobbying efforts to the contrary, first, there is little about the business models at the core of the phenomenon which would justify altruistic notion such as sharing or helping out: the ondemand economy is clearly driven by economic incentives.Second, the language of 'gigs' and 'tasks' and 'rides' devised by different operators must not obscure the reality of the underlying arrangements, which might -as we go on to demonstrate in this section -often closely resemble the working conditions found in traditional low-wage sectors.
In charting the rise of the on-demand economy in the United Kingdom, the present section is structured as follows.We begin by presenting the best available evidence on the size of the 'on-demand economy' in the United Kingdom, before highlighting key concerns about the working conditions (in particular low pay and unpredictable work) associated with the phenomenon.We then situate on-demand economy work in the broader context of casual work in the United Kingdom, highlighting in particular the parallels and close overlap with the much-discussed phenomenon of 'zero-hours contracts'.

a) Measuring Work in the On-Demand Economy
At first glance, it is nearly impossible to determine the extent of on-demand work in the UK's broader labour market developments: official UK labour market statistics -in particular the Labour Force Survey -do not contain any specific measures concerning on-demand, 'sharing' or 'gig' economy work. 3In April 2016, the ONS released a detailed feasibility study on measuring the sharing economy, presenting several potential options and data sources, as well as a wide range of problems; 4 an updated report from September of the same year 'identified 2 overarching research questions which we will be taking forward -each with differing stakeholders and policy implications […]: what is the value of the sharing economy?[And] how is the sharing economy impacting on the labour market?' 5 Whilst first results are due to be published in August 2017, they are likely to be of limited utility; not least because of an initial focus on two sectors -property rental services, and transportation platforms, only one of which will have a significant impact on labour markets, given the former sector's asset -rather than labour-intensive business models.
In the absence of official data, it is surprisingly difficult to present clear findings as to the economic and social situation of on-demand workers in the UK -even though such information ought to form the backbone of any informed debate.Indeed, the lack of data surrounding crowd work is reflected in the on-going difficulty of measuring work in a rapidly fragmenting labour market more generally. 6The best currently available evidence can therefore only be drawn from private studies.As regards the economic situation of on-demand economy workers, first, in February 2016, the Foundation for European Progressive Studies (FEPS) and UNI Europa published a survey of crowd work in the United Kingdom which had been authored by Ursula Huws and Simon Joyce at the University of Hertfordshire, in collaboration with Ipsos MORI. 7Both the headline figures and the more fine-grained picture contained in this survey are of significant interest for present purposes.
On the basis of a representative online survey of 2,238 adults aged 16-75, the authors suggested that over 10% of the working-age population in the United Kingdom have worked for a gig-economy platform -the equivalent of approximately 4,9 million workers. 8Whilst crowd work spans across all age groups, young workers are particularly heavily represented, with the 25-35 years-old age cohort making up a third of on-demand economy workers alone. 9Interestingly enough, students represented but a comparatively small share of workers, with a mere 6% regularly engaging in platform-based work.On-demand economy workers' earnings tend to be low: over 40% earn a gross annual income of less than £20,000;11 a figure which is particularly troubling when considering the fact that crowd workers will usually bear the vast majority of cost associated with their work -from providing tools such as their cars or computers to paying for upkeep and running cost.The trade union GMB, for example, has calculated that even top-rated and -paid Uber drivers in London will earn no more than £5.68 per hour once their cost (including for example car rental or depreciation, platform commission, fuel, and ancillary services) are taken into account. 12As of April 2017, the UK's adult minimum wage (for those aged 25 and over) will be set at £7.50; it is due to rise further to exceed £9 per hour by 2020.There are, however, increasing concerns that on-demand economy workers will not always benefit from this increase given their (supposed) self-employment status (on which see more in the legal discussion, below). 13This would not necessarily be a problem, if on-demand work was an occasional occupation undertaken to supplement a stable primary income.Crucially, however, it appears that such arrangements have become the only, or at least main, income source for a substantial number (up to one third) of crowd workers in the United Kingdom. 14The fact that many workers thus find themselves 'piecing together a livelihood from a range of different tasks' is reflected in the fact that over 65% are registered on more than one platform. 15he social situation of on-demand economy workers is heavily influenced by this precarious economic reality: whilst there is a small group of highly successful 'micro-entrepreneurs' who have been able to harness the potential of gig-economy work to their advantages, the reality for the vast majority is very different.As Huws and Joyce suggest, there is 'a wide variety of work.The range is extremely broad, from high-skill professional work at the one extreme to running errands at the other.' 16 It is not surprising therefore that the 'rewards of new forms of employment contract accrue to a minority, while others lose out.[…] Benefits such as pensions and sick leave are often attached to permanent jobs and increase with longevity.As jobs fracture, individuals who switch jobs, work as consultants or run "micro-businesses" with one or two employees need similar support.' 17 Given the current legal regime in the United Kingdom, however, it is not necessarily clear how such support could be achieved; a question to which discussion will return in sections two and three.

b) On-Demand Business Models and Casual (Zero-Hours) Work
Having set out the contours of on-demand work in the UK labour market, it is important to note a fundamental point: from an employment law perspective there is little, if anything, that is novel about the work arrangements underlying different on demand platforms' business models.Instead, they closely mirror working conditions in casual work settings more broadly -particularly when compared with zero-hours work, one of the most high-profile labour market issues in the UK over the course of the past years: in most contexts, on-demand economy work can be characterized as a set of digitally intermediated zero-hours contracts. 18ne of the most difficult challenges in defining Zero-Hours Contracts is the fact that despite many assertions to the contrary, they are not a single category of work arrangements.Instead, they stand for a wide range of arrangements in which workers are not guaranteed a fixed (or indeed any) number of hours in a particular period and, at least in theory, in which they are not obliged to accept any offers of work which might have been made by their employer.
Despite this wide factual variety, there is a frequent assumption that clear-cut definitions could be found -as is now the case with a newly-inserted section 27A of the Employment Rights Act of 1996 ('ERA 1996'), sub-sections (1) and 92) of which stipulate that (1) In this section "zero hours contract" means a contract of employment or other worker's contract under which: (a) the undertaking to do or perform work or services is an undertaking to do so conditionally on the employer making work or services available to the worker, and (b) there is no certainty that any such work or services will be made available to the worker.
(2) For this purpose, an employer makes work or services available to a worker if the employer requests or requires the worker to do the work or perform the services.
The main problem with such accounts is their underlying assumption that the Zero-Hours Contract is a unitary category -an assumption difficult to sustain in the face of widespread factual complexity. 19t is important to note that despite the new statutory definition, there can be no clear divisions or watertight categories in precarious work arrangements. 20The 'Zero-Hours Contract' label should not be seen as representing a clear or overarching category or organising principle of precarious work, or as somehow cleanly mapping onto received understandings of 'atypical work'.There is a considerable degree of 'heterogeneity of temporary work', 21 reflected in 'a growing nomenclature of 'atypical' and 'non-standard' work, apart from commonly used categories such as temporary, part-time and selfemployed work.Terms include 'reservist'; 'on-call', and 'as and when' contracts; 'regular casuals'; 'key-time' workers; 'min-max' and 'zero-hours' contracts.' 22Indeed, the various categories of 'atypical' work can frequently overlap, for example where agency work incorporates a 'zero-hours contract dimension'. 23As Kilner Brown J in Mailway suggested, the claimant's Zero-Hours arrangement there 'in one sense […] was casual labour; in another sense it was part-time work.' 24 Zero-Hours arrangements thus represent various degrees of fragmentation of work -from reasonably regular and consistent employment to a spot-market in labour. 25Rather than forming a single or unitary category, they represent some of the many possible variations of employment, ranging from 'preferred choices, well-paid and secure' 26 to 'vulnerable' 27 or 'poor work'. 28The use of algorithms to effect sometimes extremely short-term zero-hours contracts might thus be the most extreme instantiation of the phenomenon (yet); but it should not be seen as in any meaningful way analytically or legally distinct.
In concluding our argument that on-demand economy work closely resembles existing zero-hours models, it is also important to highlight the wide-spread (and misguided) perception that Zero-Hours work arrangements are a relatively new phenomenon, rather than part of a much larger 'tendency toward numerical flexibility [which has been] particularly marked [since] the 1980s'. 29Litigation arising from the use of Zero-Hours Contracts to allow employers numerical flexibility and to attempt to avoid the application of statutory protection can be traced back nearly forty years. 30In Mailway, 31 for example, the claimant postal packer 'could and would only attend for work in accordance with the need expressed by the employers.' 32A study by Katherine Cave in the 1990s showed the already widespread use of 'something that could be classified as zero hours contracts'; 33 with a strong growth trend as an area where there has been abuse' 34 continuing in the subsequent decade. 35The technology underpinning work in the on-demand economy is new: the underlying reality of precarious work is not.
This point can be further illustrated by reference to a recent UK Employment Tribunal ruling that Uber drivers were workers employed by the company and thus entitled to minimum wage and other basic labour rights.In Aslam v Uber BV, the court focused on the 'practical reality' of the relationship between Uber and its drivers and showed that the drivers, regardless of contractual statements to the contrary, are to be qualified workers.The mechanisms designed by Uber to insist that its drivers were self-employed merited a high degree of 'scepticism': 'Armies of lawyers' could not help a business avoid employment liabilities, despite contractual 'fictions, twisted language and even brand new terminology'. 36he employment tribunal's ruling was widely welcomed by commentators.The tribunal's focus on the reality of the relationship between the parties and its clear and accessible analysis of the law as well as Uber's business are a role model for similar cases pending in jurisdictions worldwide.The basic protection now afforded to drivers will overcome some of the worst problems-not least, because the tribunal (rightly) ruled that a driver is 'working' for the entire time that his (the vast majority being male, as noted in the decision) Uber drivers' app is switched on, and he is able and willing to accept rides.In the longer run, however, Uber drivers -even when classified as workers -will still face many of the problems encountered by zero-hours workers across the United Kingdom: from low income to struggling with unpredictable shifts due to a lack of guaranteed work.Restoring the scope of employment protection is but the first step on the road to decent work in the on-demand economy.

The Taxonomy of Personal Work Relations in the United Kingdom: from a Binary Divide to a Tri-or Multi-Partite Regime
Having in the previous section presented an outline of the main factual and legal features of the 'on-demand economy' in the United Kingdom, we now proceed to describe in somewhat more detail the creation and evolution of the very complex legal taxonomy of personal work relations within which that economy has developed and upon which it to a significant extent depends.It will be argued that the key feature of that taxonomy is the way that it has taken shape as an intricate and uneasy combination of two different classification systems, one of which is bi-partite and the other of which is tripartite in 29 S Fredman, 'Labour Law in Flux: The Changing Composition of the Workforce' (1997) 26 ILJ 337, 339.The effects were originally particularly marked in the case of female workers (339-340) and industries such as construction and dock-working (340), or trawler working.See also P Leighton and R Painter, '"Task" and "Global" Contracts of Employment' (1986) 15 ILJ 127; A McColgan, Just Wages for Women (OUP 1997) 391. 30In the instance cited, the employer's minimum guarantee payment for employees within the meaning of s character.It is that ambiguity between bi-partite and tri-partite classification systems which has come to have a crucial bearing upon the application of employment law to the on-demand economy.This is a problem which has developed incrementally, in a sequence of law-making activity upon which it will be useful briefly to recapitulate.During the period from the early 1960s to the middle and later 1970s when the main building blocks of the United Kingdom's system of statutory employment rights were being put in place, a straightforwardly bi-partite contractual classification of personal work relations was applied to designate the relational boundaries, or the 'personal scope' of the new employment rights which were being created.Closely following the contours of the tax and social security systems, the legislators made and the courts applied a binary divide between 'employees' who were engaged under contracts of employment and the 'self-employed' or 'independent contractors' operating under 'contracts for services'. 37Most of the new employment rights which were created at that period applied only to employees and not to the self-employed.For example, the right not to be unfairly dismissed, originally created by the Industrial Relations Act 1971 applied only to employees with a stated period of continuous contractual employment with the employer, originally set at two years, a qualification which has been varied from time to time but which applies today. 38here were significant exceptions to this rule, in the form of employment rights which applied on both sides of this binary divide.This was in a sense true of the rights and obligations in respect of health and safety at work which were created by the Health and Safety at Work etc Act of 1974. 39.Moreover, this was also true, in intention at least, of the big wave of employment equality or employment discrimination legislation which swept in from 1970 onwards in the shape first of the Equal Pay Act of that year, then of the Sex Discrimination Act 1975 and the Race Relations Act 1976, to be followed by the Disability Discrimination Act of 1996 and quite a host of subsequent measures. 40These measures applied to 'employment under a contract of employment, a contract of apprenticeship or a contract personally to do work' as the current consolidated version of the legislation, the Equality Act 2010, now puts it 41 .This was a formulation which appeared in a number of variants which in our view were intended to apply to all personal work contracts, whether of employment or for self-employment, and which were so interpreted by the courts over many years subsequent to their original enactment. 42his bi-partite system, under which employees enjoyed a full set of employment rights while the personally self-employed were confined to a much more limited set of such rights, though for a long time apparently a robust one, had by the later 1990s become fragile under the pressure of a growing perception that the 'employee' category was an unduly exclusive one, 43 especially in the face of the rapid growth of 'marginal' or 'atypical' forms of employment, a set of developments which now appears as a precursor of the emergence of the 'on-demand economy'.Particular strains were placed upon the binary divide between employment and self-employment as employing enterprises began vertically to disintegrate. 44Once largely released from the constraints of a previously vigorous collective bargaining system, business undertakings increasingly chose to meet their labour requirements by means of casual forms of engagement, such as those of temporary agency employment, and moreover as they increasingly cast the contracts for such engagements in the mould of self-employment.The resulting pressure points on the bipartite system became more acute as the courts, so far from relaxing the tests for employee status to accommodate casual forms of working within the category, actually restricted that category by imposing tight qualifications of 'continuing mutual obligation' for the employer to provide work and for the employee to perform work when called upon, and moreover to do so strictly in person rather than by the occasional substitution of others. 45This was an approach to the construction of the employee category which would increasingly enable employing enterprises to formulate work contracts so as to take the workers in question out of that category both by designating them as self-employed and by inserting 'no obligation' and 'substitution' clauses into the contracts to confirm that designation 46 .
Towards the end of the 1990s, the New Labour government then in power implemented a legislative strategy designed partially to mitigate these exclusionary effects of the operation of the employee category: a new and more extensive personal work status, that of 'worker', was devised and deployed as the qualifying category for the important new employment rights which were introduced by the National Minimum Wage Act 1998 and by the Working Time Regulations of 1998, implementing the EU Working Time Directive of 1993 which was then in force 47 .This new contractual category was defined in an elaborate formulation which has been embodied in section 230 of the Employment Rights Act 1996; it is as follows: '(3) In this Act "worker" […] means an individual who has entered into or works under (or, where the employment has ceased, worked under), (a) a contract of employment, or (b) any other contract, whether express or implied and (if it is express) whether oral or in writing, 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; and any reference to a worker's contract shall be construed accordingly.' It was this rather complex formulation which eventually super-imposed a firmly tri-partite system of classification of personal work contracts upon the previously bipartite one which we have hitherto described; we proceed to explain that transition in more detail.
It will be apparent that the definition introduced by section 230(3) has two parts or limbs, (a) and (b), and hence it has become customary to speak of 'limb (a) workers' and 'limb (b) workers' to designate the two types of contractual sub-categories which between them compose this new 'worker' concept or status.It is our view, though subsequently as we shall see this was to become a matter of controversy, that this new composite category was intended to be and was logically set up to be an intermediate third category interposed between the two previously existing categories, so that it represented a tri-partite system of classification of contractual work statuses. 48As we have seen, the existing bi-partite system recognised the two categories of, on the one hand, employment as an employee under a contract of employment, and, on the other hand self-employment under any other kind of personal work contract.This new third category straddled the two existing categories, being larger than either one but smaller than both the existing categories put together.It was larger than either of the existing categories in that it comprised the whole of the existing employee category in limb (a) and also included some of the existing self-employment category in limb (b); but it was smaller than both existing categories put together in that limb (b) did not include all contracts for personal self-employment, since it carved out and excluded an excepted category where the status of the counterparty to the contract was 'that of a client or customer of any profession or business undertaking carried on by the individual [who under- 45 This mode of judicial testing for employee status had been largely initiated by the Court of Appeal in O'Kelly v Trust House Forte plc [1983]  eMPloyees, workers and the 'sharing econoMy' changing Practices... Mark Freedland -JereMias Prassl takes the work]'.The intention underlying that exception seems clearly to have been that of excluding genuinely independent contracting for personal work, thereby leaving the composite category of worker to include not only employees as traditionally defined but also those working under semi-dependent or economically dependent personal work contracts.
These legislative initiatives could perhaps have been seen as promising to create a clear and coherent tri-partite taxonomy of personal work contracts and relations; but any such promise was not realised, and instead it gave rise to a classification regime which was both mixed and confused as between bi-partite and tri-partite taxonomies.The new classification regime was mixed in the sense that the new tri-partite taxonomy for employment had to co-exist with the still resolutely bi-partite taxonomy which applied in tax law and practice and social security law and practice -this is a point to which we revert in the concluding section of this paper.This alone has been a source of continuing confusion; but we suggest there has been a deeper confusion within the fold of employment law itself, since even within that domain there has been an enduring uncertainty as to how the 'worker' concept which typifies the new tri-partite taxonomy maps onto the two categories of the traditional bi-partite classification.In particular, the problem has been to work out how far limb (b) enlarges upon limb (a), or in other words, how large the zone of non-employee worker status should be deemed to be.
For more than a decade after the legislative initiative of 1998, this problem of how to reconcile the bi-partite and the tri-partite taxonomies, and how to apply both of them in the growing sphere of 'marginal' or 'atypical' employment was a perceptible one but not an evidently urgent one.It seemed that limb (b) of the worker category might not extend very far beyond limb (a); for example, the very influential decision of the Employment Appeal Tribunal in the case of Byrne Brothers (Formwork) Ltd v. Baird & Ors 49 suggested that a rather similarly strict requirement of mutual obligation might apply when it held that 'in considering the latter question the boundary is pushed further in the putative employee's favour', so that the distinction between employee status and worker status was best understood as one of degree, rather than kind. 50However, this was not necessarily an acute problem if the courts were prepared to take an expansive view of the scope of limb (a); and the Supreme Court famously did precisely that in 2011 in the leading case of Autoclenz Ltd v Belcher & Ors, 51 in which the personal work contract of a 'zero-hours' car-wash valet was held to be a contract of employment falling within limb (a) despite the best efforts of the employing enterprise to present it as a contract for self-employment falling outside both limbs of the 'worker' definition.
However, the problem of reconciliation of the bi-partite and tri-partite taxonomies was in truth becoming a more pressing one, as it became more and more important to respond to the rapid development of 'zero-hours contracts' and of the 'on-demand economy' more generally as described it in the previous section of this paper.Moreover, that rather complex mixed taxonomy was to experience a sudden, unexpected, and rather curiously deep legal disruption.
On the very same day on which the Autoclenz case was decided by the Supreme Court, judgment was also given by the Supreme Court in the case of Jivraj v Hashwani. 52The latter decision was at first glance very differently located in a dispute concerning the appointment of a legal arbitrator under a large commercial contract but turned out upon closer inspection to have profound implications for the taxonomy of personal work contracts not least in the context of the emerging 'on-demand economy'. 53his decision reinterpreted and re-configured the category of 'employment under a contract of employment, a contract of apprenticeship or a contract personally to do work' which was identified earlier in this paper as the one which is used as the qualifier for employment equality rights, and which used to be regarded as encapsulating the whole range of personal work contracts on both sides of the binary divide, and therefore as specifying and including both the two contract types which were separated by that divide.The decision was in essence that the words 'employment under' confined this whole employment 49 [2001] UKEAT 542_01_1809, [2002] ICR 667. 50Ibid at para [17] (5), at page 678H (per Recorder Underhill QC). 51 eMPloyees, workers and the 'sharing econoMy' changing Practices... Mark Freedland -JereMias Prassl equality category to those personal work contracts under which the employed person was in a relation of subordination to the employer -so that the relation or contract with the legal arbitrator which was at issue in the Jivraj case itself fell outside the scope of the employment equality legislation.
The courts and the commentators have been grappling since that time with the resulting disruption to the whole system of taxonomy of personal work relations and contracts of which this employment equality category turned out to be an unexpectedly crucial component.We suggest that this disruption can best be understood as having significantly aggravated the tension between the bi-partite and tri-partite classification systems, and added to the confusion around that central ambiguity.Clearly, the result of the Jivraj decision was that the outer envelope of the whole category of personal work relations which qualify for statutory employment rights was being forced inwards by the contraction of the employment equality category.The much more difficult question was how far this implosion was meant to go.Was it meant simply to collapse the employment equality category into the 'worker' category, while still following the contours of the tri-partite classification system?Or would this negative and reductionist impulse press on still further to carry both the employment equality category and the worker category back into a newly narrow subordinate employee category, thus restoring the bi-partite system in its full exclusionary rigour?The courts circled round this perplexing set of problems in a sequence of cases, 54 and it was not until some very recent case-law emerging from the context of the 'on-demand economy' that any significant resolution of them began to occur.This case-law and the partial resolution which they produced are presented and analysed in the following section of the paper.

The non-employee worker in the on-demand economy
In the course of recent months, the theoretically very interesting but decidedly arcane body of case-law concerning the legal taxonomy of personal work contracts and personal work relations has been electrified and illuminated by three high-profile legal decisions, two of them from Employment Tribunals and the third and most recent from the Court of Appeal.These are the cases of Uber 55 , Citysprint 56 , and Pimlico Plumbers 57 : they have attracted considerable public interest because they concern three enterprises functioning in the 'on-demand economy', one of which namely the Uber digital platform has become emblematic of that economy; we argue that their cumulative impact upon employment law's taxonomy of contracts and relations has been a very significant one.Between them they have considerably reinforced the tri-partite classification of personal work contracts, somewhat transforming it in the process, and leaving in their wake a number of complexities and some possible concerns with regard to those ultra-casual personal work arrangements known as zero-hours contracts of employment which are, as we have seen, so widely deployed in the on-demand economy as to have become almost typical of it.
The detailed accounts, as provided in the respective judgments, of the personal work arrangements which were under consideration in these three cases -those of Uber-drivers, of a cycle courier, and of an on-call plumber -between them provide a good representative sample of the wide spread and variety of work relations which are loosely comprehended within the 'on-demand economy' as generally described in the first section of this paper.In all these three cases, the claimants succeeded in showing that they were 'workers' within the meaning of Section 230(3) of the Employment Rights Act 1996 so as to be entitled to national minimum wage rights and working time rights; and in Pimlico Plumbers the claimant was also held to be qualified for employment equality rights under Section 83(2) of the Equality Act 2010 -that is to say, it was held that his work relation fell within the category of 'employment under a contract of employment […] or a contract personally to do work', the category which had been re-configured in the Jivraj case.It is not surprising that each of these three cases was perceived by the news media as a victory for the workers concerned, or that the three cases were regarded as cumulatively signalling a major advance of employment rights into the domain of the on-demand economy.This paper argues that those are largely correct perceptions, but that some notes of caution nevertheless need to be sounded.
On the positive side, from the perspective of the development of employment rights in the ondemand economy, it can be said that these three decisions do jointly and severally reinforce a reasonably expansive construction of the 'worker' category, extending its advance into the on-demand economy, as initiated by the Autoclenz case, across quite a broad front.The fact that the work relations in each of these three cases was regarded as falling within limb (b) of the 'worker' definition marks out limb (b) as unequivocally more extensive and therefore inclusive than limb (a); it seems to confirm in particular that there is a less stringent requirement of mutual obligation for personal work contracts to be admitted under limb (b) than applies under limb (a), or perhaps indeed that mutuality of obligation is not a relevant element to be taken into account in determining worker status under limb (b), as certain recent cases had begun (erroneously, in our view) to suggest. 58he decisions furthermore clearly delineate a territory within which a work arrangement may mark out the working person as a 'worker' although he or she is at the same time classified as self-employed rather than being an employee and as such falling under limb (a).The idea is thus confirmed 59 that a person may, for the purposes of this particular taxonomy at least, be self-employed though working for another, thus distinguishing between persons self-employed and working on their own account (who still fall outside limb (b)), and persons self-employed but not working on their own account, for whom there is therefore a distinct conceptual basis for their admission into limb (b) of the category of 'worker'.
There was also a further positive development of employment rights in the Pimlico Plumbers decision in particular.In that case, the claim was not only for minimum wage and working time rights as a 'worker' within section 230(3) of the Employment Rights Act definition, but also for employment equality rights (in respect of alleged disability discrimination) under the 'employment' category as defined by section 83(2) of the Equality Act 2010.The Court of Appeal in Pimlico Plumbers confirmed the indication, 60 first given by Lady Hale in the Supreme Court in Bates van Winkelhof, 61 that the two categories were to be regarded as essentially convergent ones, both invoking a broad notion of 'subordination' which was implicit in EU law's notion of the worker.This means that the employment equality category of 'employment' replicates the extent of limb (b) of the 'worker' as well as that of limb (a).The two categories are now in effect being treated as identical to each other.This equating of the two classes of personal work contract or relation, the 'worker' class and the employment equality class, when coupled with the Court of Appeal's expansive approach to limb (b) in the Pimlico Plumbers case itself, could and hopefully will serve to dispel the fear, to which we adverted earlier, of a contrary movement pushing the whole combined category inwards towards a narrow conception of dependent employment almost coincident with limb (a) and rendering limb (b) virtually or entirely empty of content.
However, perhaps three caveats need to be entered, enjoining caution against regarding this recent case-law as a straightforwardly progressive development of employment rights in the domain of the on-demand economy: one is a slight cautionary note and the other is a more substantial one.The slight cautionary note is that there is still the possibility that the courts could revert to a more rigorous deployment of the requirement of mutual obligation as a restriction upon the scope of this emergent composite category of the 'worker' who is in 'employment' for equality law purposes.The finding that there was sufficient mutual obligation to satisfy that definition on the facts of Pimlico Plumbers, and the indication given by Underhill LJ that the approach taken to those facts might be extensible to a wider eMPloyees, workers and the 'sharing econoMy' changing Practices... Mark Freedland -JereMias Prassl range of cases, 62 are between them encouraging against that possibility, but it might nevertheless still present itself in relation to some of the ultra-casual 'zero-hours' work contracts or arrangements which, as we have seen, are sometimes to be found in the sphere of the on-demand economy.We would regard any such development not only as deeply problematic in practice, given its potential to leave increasing swathes of those most thought to require recourse to at least the basic standards such as working time and minimum wage protection beyond the scope of employment and anti-discrimination law, but also as indefensible in principle, as so narrow a construction of limb (b) would clearly go against Parliament's intention to extend a floor of rights beyond the traditional core category of employees.
The second caveat is a less specific one, but perhaps a more substantial one.The price of the claimants' victories in these recent cases was that they were found to be, indeed in a sense were left remaining in, the general category of self-employment in which their personal work contracts had located them, albeit in a part of that category in which they were qualified for minimum wage rights, working time rights, and employment equality rights.That was all that the claimants needed to claim, and that was all that the tribunals or courts needed to find in order to grant them their victories.But those victories were therefore only partial ones, in the sense that the workers in question were not accorded full 'employee' status under limb (a) of the definition of 'worker', and these findings perhaps encourage the view that further claims to rights dependent on that full employee status could not have been made, or if made would not have been upheld by the appellate courts.This may represent something of a retreat from the robust approach taken by the Supreme Court in the Autoclenz case, where the claimants were deemed to fall under limb (a) of the definition of the 'worker', as being employed under contracts of employment strictly so called, although a finding that they fell under limb (b) might have sufficed to sustain the rejection of the employer's appeal.
The third caveat, finally, relates to the broader labour market impact of these recent decisions. 63The employment tribunal's ruling in Uber, for example, was widely welcomed by commentators as a progressive decision in favour of workers' rights. 64The tribunal's focus on the reality of the relationship between the parties and its clear and accessible analysis of the law on worker status (as well as Uber's business more generally) will serve as a role model for similar cases pending against on-demand economy employers in the UK and further afield.An important note of caution should nonetheless be sounded: the outcome for workers might be less positive than appears at first sight.This is not, of course, because employment rights somehow mean less flexibility, as the company attempted to argue in the wake of the decision: it is perfectly possible to employ workers on decent, yet flexible terms.The problem arises from the fact that only a limited set of core rights are attached to worker status, including the National Minimum Wage Act 1998 and the Working Time Regulations 1998.Such basic protection will overcome some of the worst problems faced by Uber drivers and other on-demand workers -not least, because the tribunal (rightly) ruled on the facts that the drivers were 'working' for the entire time that their Uber drivers' app was switched on, they were in the territory in which they were licensed to work, and they were able and willing to accept rides, not just when transporting passengers in their car.In the longer run, however, Uber drivers -even when classified as workers -will face many of the problems encountered by zero-hours workers across the United Kingdom: from low income to struggling with unpredictable shifts due to a lack of guaranteed work. 65o the upshot of the recent litigation on UK worker status is that workers in the on-demand economy are tending to be placed into a slightly shaky intermediate category between full employment and full self-employment, treated as self-employed workers but not working on their own account.They seem likely to find themselves on the right side of the fence for the rights of 'workers' and for employ-62 See in particular Underhill LJ, [2017] EWCA Civ 51 at para 145:-'If the position were that in practice the putative employee/worker was regularly offered and regularly accepted work from the same employer, so that he or she worked pretty well continuously, that might weigh in favour of a conclusion that while working he or she had (at least) worker status, even if the contract clearly (and genuinely) provided that there was no legal obligation either way in between the periods of work.' 63 For further development of this critique, see R Hunter and J Prassl at http://ohrh.law.ox.ac.uk/worker-status-for-appdrivers-uber-rated/.
eMPloyees, workers and the 'sharing econoMy' changing Practices... Mark Freedland -JereMias Prassl ment equality rights, but may, if our fears about the potential (mis-) development of the law were to be realised, find themselves on the wrong side of the fence from the rights of employees with contracts of employment.In the latter case, workers might thus be liable to be in effect partly within the fold for the tri-partite classification of employment rights, but out in the cold so far as the system has remained a bi-partite one.This is a rather confused state of affairs, especially when it has to be reconciled with the tax and social security classification systems which still maintain a basic binary divide, and does not distinguish between those who are self-employed and operating 'on their own account' and, on the other hand, those who are self-employed but not operating on their own account because they are seen as working for an employer.It is this rather convoluted set of combinations of statuses which is the subject of our brief conclusion to this paper.

Worker Status Beyond Employment and Anti-Discrimination Law: Coherence and Confusion
As we hope to have demonstrated, the rise of the on-demand economy in the UK is but the latest chapter in a story dating back to the mid-1960s, in which we find the binary divide between employment and self-employment becoming gradually more contested and fragile, as the incentives and opportunities multiplied for employing enterprises to re-cast employment contracts as contracts for services, and to identify their workers as being 'self-employed', in order to minimise the application of worker-protective employment legislation, and to obtain fiscal and social security advantages. 66Crucially, however, certain incentives towards such (re-)classification of individual employment relationships could also be found on the workers' side, with independent contractor status appearing, at least in the short term, to be of financial advantage to both parties.
This final consideration arises from the legal status of on-demand economy workers as self-employed individuals, particularly in the context of tax law.In the United Kingdom, self-employed workers are taxed (for national insurance purposes) at a lower rate than the employed; a rate which can be reduced further when services are offered through the worker's own limited liability company. 67The tax authorities (HMRC) have become worried about the 'unfair manipulation' of the applicable rules (IR35), which they suggest 'contribute to, and reinforce, the incentives for individuals to work through companies when they would otherwise be employees, and for businesses to engage individuals through companies, even if they are working in a similar way to their employees' -and thus a considerable loss of approximately £430m per annum to the Exchequer.' 68his development has increasingly troubled policy makers, but hitherto there have been no comparative legislative initiatives to introduce a concept of 'the worker' designed to be more inclusive than that of 'the employee' in tax and social security law, or judicial initiatives to be more aggressive in detecting and countering employers' attempts to disguise relations and contracts of employment as those of self-employment. 69The binary divide has thus survived more or less intact in the realm of taxation and social security; indeed, in very recent years both the legislature and the courts have if anything somewhat relaxed their pursuit of 'disguised employment'. 70Certain elements of Government policy have even favoured the growth and development of self-employment, because it enables the claim to be made that the level of unemployment is constantly falling, and it reduces the cost of social security provision. 71 [2011] UKSC 41, [2011] ICR 1157. 52[2011] UKSC 40, [2011] ICR 1004. 53The discussion of this case in the ensuing paragraphs builds upon the journal article by Mark Freedland and Nicola Kountouris, 'Emploment Equality and Personal Work Relations -A Critique of Jivraj v Hashwani' (2012) 41 Industrial Law Journal 1-11.