APPLYING RIGHTS ACCORDING TO THE INSTITUTIONAL MODEL

This section begins by setting out a number of labour rights that flow from an application of the institutional approach. At the outset, it is important to reiterate that I do not propose to replace existing labour codes with a set of constitutional rights that will be subject to direct, horizontal litigation between workers and employers.

Indeed, scholars such as Langille see this as the primary manner in which constitutional rights (such as the freedom of association and the right to equal­ity) should interact with statutory labour codes.^[579] I do not disagree; that is not to say, however, that the institutional approach never goes beyond the bounda­ries of legislation. As we shall see, in some cases, such as the scope of collective bargaining, the institutional approach does do so.

It is also important to keep in mind that - as pointed out in the preface to part II - the arguments below will be shaped by the constitutional background: a constitution could have specifically articulated horizontal rights; it could have

Applying Rights According to the Institutional Model 159 a general horizontal rights clause that empowers courts to decide which rights to apply it to, and when; or it could restrict courts to interpreting and develop­ing only existing statutory law.

Let us now proceed to the discussion. Once it is determined that one private party exercises institutional power over another in a manner that can enable the former to violate the rights of the latter, the institutional approach then asks: which are the rights that are vulnerable to being infringed by virtue of superior institutional power? Scholars of horizontality will recognise that this is a formula often used by courts: in determining when a particular right applies horizontally, courts have often asked whether the duty bearer is in a position to threaten the exercise of the said right. In the specific context of employment, Collins observes that:

In exercising their disciplinary power over employees, for instance, should we not require private employers to respect civil liberties, for their disciplinary powers pose a threat to freedoms such as freedom of speech and association that seems to be as equally grave as that presented by a powerful state?^[580]

While the institutional approach eschews direct analogies with the state, the point is well taken. Now, more specifically, in the case of platform work (and the labour market more generally), as indicated above, these rights apply both at the point of entry into the work relationship and through the work relationship. As Collins points out, borrowing from Kahn-Freund's vocabulary, the former can be understood as submission, while the latter is subordination:

Submission occurs when a person enters a contract of employment on terms dictated entirely or almost entirely by the employer and where there may be no reasonable alternatives to earn an income but to take this job... the concept of subordination differs from submission, because it concerns the daily experience of an employee being subject to the hierarchical control of the employer or manager.^[581]

In more specific terms:

First, workers face capitalists within the sphere of exchange, notably in the labour market, to negotiate the terms of labour contract (stating, for example, workers' salary).

The rights in question, therefore, would include collective rights, which aim to mitigate the institutional imbalance of power so that both parties are in a

position to shape the terms of their relationship. They would also include indi­vidual rights that constitute a threshold beneath which these terms cannot go.

On the individual front, the institutional approach takes familiar vertically enforceable rights found in most constitutions, and applies them in a horizontal framework. The first is the right to a minimum wage, which flows from a right that is often found in constitutions: the right against forced labour.

The justification for holding that the right to a minimum wage is a subset of a horizontally applicable right against forced labour was set out in the previ­ous chapter, in the examination of the PUDR judgment. Here, I recapitulate the argument briefly. The minimum wage serves as a heuristic to determine situations in which it can be said that the structural constraints of the labour market essentially make workers ‘unfree’. Workers who ‘agree’ to a wage that goes below even the minimum wage (which is assumed to be the threshold remu­neration allowing for a basic, dignified life) are deemed to have been coerced or forced into the agreement, borne out of the differences in institutional power between employers and workers. Thus, as Collins et al frame the argument:

workers’ inequality of power, understood in this way, leads them to accept wages that they would not accept if the risks were distributed equally. Capitalism gives the opportunity to exploit systematically workers’ aversion to risk... the right to a fair wage should be understood as a right to address the risk imbalance that is inherent in capitalism.^[583]

The extension of the right against forced labour to the horizontal context, there­fore, translates into an enforceable right to a minimum wage (and, indeed, can arguably be extended to working hours limits and holiday pay, which cleave to the same logic).

Indeed, in the context of platform work, scholars have pointed out how the oversupply of labour^[584] - a characteristic feature of platforms - is directly responsible for depressing wages, often to below minimum wage levels.^[585] While concerns have been raised about how this might be enforced in cases where workers switch between platforms, as ACL Davies’s scholarship on working time demonstrates, this difficulty is not insurmountable within the existing legislative framework.^[586]

When it comes to the question of remedy, once again, the PUCL judgment indicates the way forward. Many jurisdictions will have a minimum wage, encoded in a statute, that applies to at least some sets of workers or employees. Litigation, thus, will normally take the form of challenging the exclusion of certain categories of workers from minimum wage laws (as was the case in PUCL); the institutional approach, as we have seen, provides a court with good reasons to extend the coverage of those laws to those excluded from them (as opposed to having to judicially fix a minimum wage). A court could do this (as the Indian Supreme Court did) by interpreting an existing, horizontal right against forced labour to derive a horizontal right to a minimum wage; in a constitution where there is only a vertical right against forced labour but a generic horizontal rights clause, the institutional approach would give a court a reason to hold why, in this context, applying the constitutional right horizontally is appropri­ate. Finally, where there is neither, but there is a statute that makes a minimum wage dependent on employment status (a plausible scenario), the institutional approach (as discussed above and as shall be shown below) could contribute to a court finding that the putative workers before it belong to the categories that are granted statutory labour law protections.

The second right that could be applied in the horizontal framework is the right to a fair trial (also found in many constitutions, and a general facet of the rule of law).

The manner in which this could be done in adjudication - across differ­ent constitutional contexts - largely follows my discussion about the right to minimum wages above: where a direct horizontal right exists, workers could invoke it. Where there is a vertical constitutional right to a fair trial or against unfair labour practices, the institutional approach would help make the case for its horizontal application; and where there is only statute, the institutional approach would inform an assessment of which parties are entitled to invoke the statutory protections on offer.

Thirdly, and closely linked, are issues of discrimination. It was famously documented, for example, that Uber cited the ‘independent contractor’ status of its drivers to justify its refusal to comply with US anti-discrimination law, which would have required it to ‘provide wheelchair-accessible transit’.^[590] Under the institutional approach, platforms would no longer be able to avoid constitution­ally mandated non-discrimination obligations (the analysis in specific contexts would, again, be similar to the discussions around the right to minimum wages and the right against unfair labour practices).

Furthermore, staying with the issue of discrimination, as discussed above, platform ‘apps’ are proprietary, non-transparent and algorithm-based. In this context, there exists a swathe of contemporary scholarship dealing with issues around algorithmic bias and algorithmic discrimination. The central role of algorithms in creating and maintaining the institutional difference in power between platforms and workers thus makes a clear case for the extension of the rights to equal treatment, and against non-discrimination, into the horizontal domain.

These issues also extend to other areas. For example, as Rosenblatt points out, ‘through ratings systems, consumers can directly input their biases into worker evaluation systems in ways that companies cannot do on their behalf’.^[591] This has many potential consequences: if, on the basis of the institutional approach, the right to non-discrimination would be held to apply inter se between platforms and their workers, there would then be a corresponding burden upon platforms to demonstrate that the ‘apps’ do not function in a discriminatory way, or with discriminatory effects. Thus far, platforms have resisted scrutiny of such issues by invoking their proprietary rights to the ‘apps’; such an argument would no longer be available - or at least, would be severely weakened - if the workings of these ‘apps’ were subjected to the obli­gations of non-discrimination.

Indeed, this argument would not be limited to non-discrimination. It would apply wherever the ‘app’ was used to violate a putative right (including the right against unfair labour practices). Recall Schein’s description of the ‘objective, impersonal authority of Uber’s software’. Once it is understood that the ‘app’ is a crucial site of the institutional difference in power between the worker and the platform, a set of guarantees against algorithmic decision-making, as well as, crucially, the right to algorithmic transparency - which exist, for example, in EU law - can be applied horizontally, inter se, between the two parties.

Finally, there is substantial controversy over safe working conditions in plat­form work. The issue of safe working conditions may complicate the platform/ worker binary, and bring into the equation a third party: the individual who is the direct recipient of the worker’s services, through the platform. For certain kinds of platform work (such as services provided in the domestic context), the institutional approach may require a shift from the platform as the duty bearer (because the platform’s institutional power does not, in fact, place that particular right at threat) to the consumer (who, in that specific context, occu­pies the position that the platform otherwise would). This may, of course, not always be the case: as the platform continues to control the terms on which both the worker and the end user can utilise it, it retains the power to make use of its services conditional upon the fulfilment of certain obligations by the end user (these could include, for example, non-discrimination obligations as well as the health and safety obligations referred to above). Responsibility in such cases may well be joint in nature.

As indicated in the previous section, the flexibility of the institutional approach is key to accommodating the heterogeneity of platform work. By plac­ing the rights bearer at the centre of its enquiry, the institutional approach is alive to situations where, within the institutionally mediated work relationship, different private parties may be in position where their institutional loca­tion enables them to violate different rights. It is important to note that there exist similar arguments in this direction: this ‘splitting’ of rights and obliga­tions has also been suggested by Prassl and Risak (and before them, by Davies and Freedland^[592] and by Langille and Davidov^[593]), who advocate a ‘functional’ approach to various features of the employment relationship and argue for placing obligations based upon which party is performing a specific functional aspect of employment.^[594] As indicated above, the responsibility may also be joint, and ‘splitting’ the work relationship into its functional components, under the institutional approach, does not necessarily entail silos as far as rights and obligations are concerned. The institutional approach provides the underlying theoretical justification that buttresses Prassl and Risak’s functional argument, as applied to platform work (it is important to note that in certain jurisdictions, such as the UK, this is no longer a serious problem or issue).

Moving on, the institutional approach applies with at least as much force to the collective aspect of labour law, guaranteeing the right to collective bargaining (flowing from the freedom of association). The logic of this is straightforward, and has been specifically recognised, inter alia, by the Supreme Court of the USA: the right to collective bargaining is required to offset the institutional power enjoyed by employers, on account of their ownership and control over capital and the means of production.^[595]

The associational rights argument, however, goes beyond the simple ques­tion of mitigation of power imbalances at the point of entering the bargaining process, and demonstrates one domain where the institutional approach goes significantly beyond existing approaches. One of the fundamental purposes of the right to freedom of association in the vertical domain is to ensure the participation of citizens in governance (in other words, substantive democracy). The institutional approach provides us with a reason to extend that logic to the domain of ‘social institutions’.^[596] In practical terms, that translates to granting workers (through collective bargaining) an ‘opportunity to play a part in the rule-making process of the enterprise’.^[597]

This is something that has traditionally been argued to be beyond the ambit of collective bargaining, which traditionally is limited to questions of wages, working hours and so on,^[598] or to the ‘sphere of exchange... rather than... the sphere of production’.^[599] In other words, the scope of collective bargain­ing has traditionally been curtailed by invoking concepts such as ‘managerial prerogative’^[600] and entrepreneurial control. Thus,

most management rights are never challenged by collective bargaining or legislated mandates on employer behavior... under collective bargaining, ‘management has the right to determine what work shall be done; to determine what kinds of services and business activity to engage in; and to determine the techniques, tools, and equipment by which work on its behalf shall be performed’.^[601]

By contrast, the institutional approach justifies a more extensive and deeper role for collective bargaining (again, flowing from the freedom of association, and its governing logic in the vertical domain) than might otherwise be allowed. It affirms Karl Klare’s proposition that ‘those whose collective efforts make social production possible should have a decisive say in the decisions that affect the process, that they pose themselves morally and institutionally as the authors of their own destinies in the workplace’.^[602]

In the specific context of platform work, therefore, this would arguably extend to not just subjecting the working of the ‘apps’ to obligations of non­discrimination, but also subordinating the platform’s proprietary claims over the ‘apps’ to the democratic process of collective bargaining over how, when and in what manner the ‘apps’ are to be deployed in regulating work. Additionally, the institutional approach here would reinforce the claims for algorithmic trans­parency that we have discussed previously, in the non-discrimination context: as long as the functioning of the ‘app’ (including, for example, the source code) are kept confidential by the platform, it is obvious that the asymmetry of informa­tion will preclude any effective bargaining on the lines discussed above.

Thus, under the institutional approach’s understanding of the scope of the right to collective bargaining, platforms can no longer claim blanket immunity from subjecting the ‘app’ to scrutiny by invoking corporate confidentiality or competition law concerns; at the very least, the institutional approach provides strong countervailing arguments against such claims.

While arguments for greater worker voice and participation are generally made within the framework of industrial democracy, what is significant here is that the institutional approach can help constitutionalise those arguments, and substantially push the boundaries of existing law. The claims outlined above would normally be excluded from collective bargaining agreements on the grounds that they pertain to managerial prerogative. The institu­tional approach provides us with constitutional reasons to challenge any such exclusion.

Finally, it is important to note that collective bargaining can take many concrete forms, and, as the Canadian Supreme Court and the European Court of Human Rights have both correctly observed, a constitution cannot stipulate the specific regime that it must take¹⁰⁹ (for example, Article 27 of the Charter of Fundamental Rights of the European Union guarantees to workers the right ‘to information and consultation’,¹¹⁰ which is then provided a concrete form by the European Works Council Directive¹¹¹). What it can do, however, is invali­date any substantive interference or undue burden that the state may seek to place upon the right (as informed by the institutional approach, in the manner discussed above), including - as discussed in section II of this chapter, in the

Germany, Sweden, and the United States’ (1984) 8(1) Hastings International and Comparative Law Review 93.

¹⁰⁹ Swedish Engine Drivers’ Union v Sweden [1976] ECHR 2; Health Services & Support v British Columbia (n 27); indeed, as Alex Rosenblatt points out, collective bargaining efforts will need to take into account the different forms of platform work within the same platform. Rosenblatt, Uberland (n 1) 54. But see Fraser v Ontario (Attorney General) (2008) 92 OR (3d) 481 (Supreme Court of Ontario); the critique in B Langille, ‘Why Are Canadian Judges Drafting Labour Codes — and Constitutionalising the Wagner Act Model?’ (2010) 15 Canadian Labour and Employment Law Journal 101.

¹¹⁰ Charter of Fundamental Rights of the European Union (2012/C 326/02), Art 27.

¹¹¹ Directive 2009/38/EC of the European Parliament and of the Council, of 6 May 2009. account of collective laissez-faire - restrictions under existing common law.^[603] This would also extend to legal burdens, such as anti-trust and competition laws, that have been used to defeat efforts at collective bargaining, especially where rights are available only to employees but not to other workers (for example, as of 2002, more than 30 million US workers were excluded from exercising their associational rights^[604]). Thus, the advantage of the institu­tional approach, as applied horizontally, is that it precludes employers from invoking anti-trust or competition laws against the right to collective bargain­ing, and also precludes arguments predicated upon the non-employee status of workers to do so.^[605]

The role of the institutional approach in such situations can be illustrated through a set of recent examples. In R (on the application of Independent Workers Union of Great Britain) v Central Arbitration Committee and Roofoods Ltd t/a Deliveroo,^[606] the UK High Court (and subsequently the Court of Appeal) ruled that, because of the existence of a substitution clause, Deliveroo riders were not ‘workers’ within the meaning of section 296(1)(b) of the Trade Union and Labour Relations Act 1992, and were not in an ‘employment relationship’ in the context of EU law. The consequence of this, as Bogg points out, is that Deliveroo workers were deprived of their right to unionise, and Deliveroo was entitled to invoke competition law against them in case they attempted to do so, along with potential discrimination and blacklisting.^[607] Bogg argues that the disposi­tive weight given to the substitution clause by the High Court was incorrect in view of ILO Recommendations on the employment relationship, which had been incorporated into EU law by the European Court of Human Rights’ Grand Chamber.^[608] For the reasons advanced in the previous section, the institutional approach provides strong theoretical ballast to supplement Bogg’s argument that substitution clauses ought not to be dispositive in determining the question of employment status: the logic underlying the substitution clause argument - that it negates any personal obligation to perform work - has little bearing upon the institutional power differences that structure the contract. Thus, much as the institutional approach justifies giving less weight (for example) to the traditional ‘employment’ test of whether the worker can switch between different putative employers, it also justifies giving less weight to the existence of a substitution clause in determining what rights a worker ought to enjoy.

To add to this, if, hypothetically, the institutional approach was accepted as providing interpretive guidance to courts (in the manner outlined previously in this chapter), the High Court would be required to interpret the word ‘worker’ in a manner that protected Deliveroo workers’ right to association, instead of negating it (an argument Bogg himself makes, in view of the fact that Article 11 of the European Convention on Human Rights is a ‘fundamental right’). This would be done by holding that Deliveroo workers fall within the definition of ‘worker’, and it would then follow - without a court having to do more - that they do have the right to unionise, as already provided by the statute. It is in this way that, acting in lockstep with existing legislation, the institutional approach can articulate a background framework of constitutional, horizontal rights that provide the context within which labour laws can be interpreted so as to protect the rights of platform workers.

However, there are also examples where the institutional approach’s role is not towards the evolution or modification of the law, but is an explanatory one. This is equally important. For instance, in this chapter, I have discussed at some length the example of ride-sharing platforms as a central case in consider­ing the application of the institutional approach to platform work. In Uber v Aslam and Ors, the question of whether Uber drivers were ‘workers’ within the wider ‘Limb B’ definition under section 230(3)(b) of the Employment Rights Act of 1996^[609] (which goes beyond the requirement of an employment relationship and only requires that an individual ‘perform personally any work or services for another party’) came up for decision. Both the Employment Tribunal and the Employment Appellate Tribunal held that the drivers were workers. This conclusion was affirmed by the Court of Appeal, which considered 13 indicators (originally set out by the Employment Tribunal) in support of its finding.

Looking at these indicators closely, it is clear that what specifically weighed with the Court of Appeal, among other things, were: information asymme­try (and control over passenger data); Uber’s power to ‘log off’ drivers for not accepting trips; the algorithmic rating system, which ‘amounts to a performance management/disciplinary procedure’;^[610] and Uber’s power to unilaterally amend terms. It will be noticed that each of these indicators was discussed in the previ­ous section as an instance of the institutional power wielded by platforms such as Uber, which enables the violation of rights. While the Court of Appeal did not frame its analysis in terms of institutional power, it did note that in labour law, the concept of a ‘sham’ contract had evolved to take into account:

[T]he convoluted, complex and artificial contractual arrangements, no doubt formu­lated by a battery of lawyers, unilaterally drawn up and dictated by Uber to tens of thousands of drivers and passengers, not one of whom is in a position to correct or otherwise resist the contractual language.^[611]

Notably, the holding of the Court of Appeal was upheld, in relevant part, by the UK Supreme Court, applying similar reasoning. Specifically, the Supreme Court noted that, through the indicators that it had set out, the Court of Appeal had correctly demonstrated that the elements of subordination and dependency were fulfilled in this case.^[612]

I submit that, for the reasons advanced above, a close reading of the judg­ments in the Uber case reveals a shift from the interpersonal view of power to the structural or institutional view of power in the context of labour law. Indeed, the judgments can be best explained and justified from this framework of analy­sis. This is particularly evident in the 13 factors discussed by the Employment Tribunal and the Court of Appeal, which shifted away from the old indicia of ‘employment’ to indicia more suited for the decentralised, institutional power at issue between platforms and workers. While the UK courts did not, of course, call what they were doing ‘the institutional approach’ (locating it, rather, within a shift towards a more purposive understanding of the labour contract), as pointed out above, the institutional approach serves as a better explanatory basis for the judgment(s), as well as providing theoretical support for the courts’ approach, given that this is still contested terrain. According to the institutional approach, therefore, the judgments in the Uber case were rightly decided, their best justification lies in reasoning that takes into account institutional power imbalances and, finally, this form of reasoning should be extended to other cases involving the platform economy as well.

It is also important to note that an important ancillary question in the Uber case involved the question of when the drivers qualified as workers. The Employment Tribunal held that the drivers were ‘working’ from the moment that the ‘app’ was switched on, and this finding was affirmed by the Employment Appellate Tribunal, the Court of Appeal and the Supreme Court. Of particu­lar interest is the Employment Tribunal’s observation that ‘it is essential to Uber’s business to maintain a pool of drivers who can be called upon as and when a demand for driving services arises’.^[613] As noted previously in the chap­ter, the existence of a ‘reserve pool of labour’ is central to institutional power within the labour market, and particularly central to the platform work model. Consequently, the finding that drivers were ‘workers’ - and therefore entitled to Limb B rights from the time the ‘app’ was switched on - flows directly from the structure of argument set out in the previous section: it is the institutional difference in power that enables more powerful parties to violate the rights of

Applying Rights According to the Institutional Model 169 the less powerful. The purpose of the institutional approach is to accurately identify the nature of this difference in power and apply the rights framework to mitigate it.

A final, illustrative example is that of The Independent Workers’ Union of Great Britain v The Secretary of State for Work and Pensions.^[614] In this case, the UK High Court held that certain EU Directives pertaining to the health and safety of workers had not been properly transposed into domestic law by the UK, in contravention of its obligations under the EEC Treaty. The UK’s defence turned upon the distinction between ‘employees’ and ‘workers’, and the argument that EU law required it to extend those health and safety benefits only to employees and not to ‘Limb B’ workers. This argument was rejected by the Court, which then went on to find that, on merits, the UK had discharged its obligations in some respects, while in other respects it had not. Importantly, a dichotomy between the rights enjoyed by ‘employees’, which included ‘welfare’, and the rights enjoyed by Limb B workers, which were limited to ‘health’ and ‘safety’ but did not include welfare, was upheld by the Court, because the overarching EU Directive was itself limited to ‘health’ and ‘safety’. This case reveals that in jurisdictions that have introduced the three-tiered employment law structure, there remain problems with respect to differential treatment that cannot always be resolved by pure statutory construction with­out reference to overarching normative principles (as in this case).

Could the institutional approach make a difference? Potentially, yes. The institutional approach here would ask what justifies excluding Limb B workers from the same ‘welfare’ rights that are afforded to full employees (in addition to ‘health’ and ‘safety’). Where no justification can be offered, the institutional approach can become either the basis of law reform proposals or, in jurisdictions where it is possible, the basis for a narrowly tailored constitutional challenge, based on the principles of non-discrimination, that will seek to extend the rights already available to employees under legislation to the more extensive category of ‘workers’.

A concluding point: it is important to note that these are only a basic mini­mal set of constitutional rights that are designed to mitigate the violations that flow from the institutional difference in power that exists in the work relation­ship. They do not restrict what additional rights might be accorded, whether through statute or through the collective bargaining process.^[615] They thus aim at ‘collective bargaining from a base of expanding social protection’.^[616] Scholars have expressed concerns about the ‘juridification’ of labour law and how consti­tutionalisation can inadvertently serve to harm labour rights by taking issues

out of the political domain and into the realms of courts.^[617] This concern is justifiable. That is why it is important to stress that the role of the institutional approach is not to supplant or substitute for either private law or collective bargaining, but to provide for a threshold of rights that workers can resort to in the absence of statutory labour law (one of which, as we have seen, includes the right to collective bargaining).

VII.