THE INSTITUTIONAL MODEL AS APPLIED TO PLATFORM WORK

I will now examine how the institutional approach might be applied to the case of platform work. As discussed in section II of this chapter, traditional labour law is founded upon the employment relationship/independent contractor binary, which determines which set of workers are entitled to avail of the pano­ply of labour rights.

To do this, courts have evolved a set of indicators, which are remarkably simi­lar across jurisdictions. However, as pointed out above, these indicators do not always take into account the features highlighted by the institutional approach. Relevant to our purposes, for example, one significant factor across jurisdictions is whether the worker is free to work for more than one putative ‘employer’ or is contractually bound to offer their services only to a single employer. The logic of this test harks back to the ‘traditional employment relationship’, where an employee worked in a single factory or upon a single shop-floor, as opposed to the freelancer, who would perform many different contracts with different contracting partners. And as is evident, the test presupposes the personalised vision of power, located within a single, specific entity, which is precisely what the institutional approach seeks to displace; as Bogg’s observations (in the previous section) make clear, ‘decentralised domination’ is no less a form of domination.

When applied to platform work, where the existence of multiple platforms and ‘decentralised domination’ are more features than exceptions, this indica­tor is particularly inapposite. Indeed, it is reminiscent of the old ‘monopoly argument’ in deciding when to enforce rights against non-state parties.

Another indicator invoked by courts is whether the worker’s equipment is provided by the putative employer or whether they are expected (or required) to use their own equipment (such as the car in the case of ride-hailing services such as Uber).

In the case of platform work, however, a focus on equipment poten­tially diffuses attention from the source of platforms’ institutional power. As Rosenblatt observes in the context of Uber:

The company determines the type of cars that are eligible on its platform, and it sometimes modifies the list of acceptable types at will; sets and changes the pay rates as it wishes; controls the dispatch; targets drivers unevenly with its incentives; retains full power to suspend or fire drivers without recourse; and mediates and resolves conflicts at its discretion, ranging from issues of passenger disputes to wage theft. An algorithmic manager enacts its policies, penalises drivers for behaving in a manner unlike what Uber ‘suggests’, and incentivizes them to work at particular places in particular times. (emphasis added)^[561]

Central to the functioning of what Rosenblatt calls this ‘algorithmic boss’ is the ‘app’. As Vincenzo Pietrogiovanni points out:

Uber’s activity is essentially based on the applications that connect drivers and passengers. The ‘ride’ follows and it is followed by other several important steps: creating an account; reserving the ride; rating the driver; checking information relat­ing to drivers (eg their driving licence or the condition of their cars) - all of which is completely owned and directly managed by Uber, without any participation or direct access to the process by the drivers.

Similarly, Prassl and Risak point out that:

The platform is fully in control of the existence of the relationship. Drivers have to sign up with Uber to use its app; but the process is significantly more intensive than mere registration: the company will vet drivers’ cars, check their licences and demand proof of relevant insurance for the jurisdiction in question. Uber is similarly in charge of terminating its driver’s access to the platform.^[563]

Prassl and Risak further point out that the ‘ratings system’ employed by the platform is used as a gateway for access to the ‘app’ (workers whose ratings slip below a certain level can be ‘booted’ from the ‘app’) and, in some cases, provides privileged access within the ‘app’.^[564] This, then, effectively serves as a form of performance appraisal^[565] (with consumers of the service playing the role of ‘middle managers’^[566]).

It is therefore clear that the institutional difference of power between the platform and the worker is sustained and entrenched by the ‘app’, which is exclu­sively owned and controlled by the platform. Depending upon the platform in question, the ‘app’: (i) serves as a gateway to entering the work relationship;

(ii) is used to surveil and control behaviour during the course of the work rela­tionship (for example, determined routes in the case of ride-sharing services);

(iii) is used to determine remuneration for work (through algorithms that, again, are designed by the platform), including the unilateral slashing of rates, which workers must accept in order to keep accessing the platform;^[567] and

(iv) through its ratings system, is used to discipline and exclude workers.

Control and ownership of the ‘app’, therefore, serves the analogous purpose that control of capital and the means of production did in the traditional labour market, which we discussed in the previous sections: ‘the labour process is embedded in the company’s proprietary software, and the company programs the software to design, direct, and surveil the worker’.^[570] Control and ownership over the equipment used for the work, by comparison, pales into insignificance (as Brishen Rogers puts it, the ‘relative investment’ of the platform remains of a far greater order, because of the ‘app’^[571]). Rebecca Schein puts the point most eloquently, and in the language of the institutional approach, when, in the context of disputes over working time, she notes that ‘drivers who dispute the company’s time-keeping accuracy are confronted - via ostensibly powerless

The Institutional Model as Applied to Platform Work 155 service reps - with the objective, impersonal authority of Uber’s software’ (emphasis added).^[572]

At the same time, however, as pointed out at the beginning of this chapter, platform work is of different kinds, and there exist situations where the plat­form itself genuinely serves as little more than as an intermediary facilitating workers to enter into independent contracts with those in need of their services. This includes a range of circumstances where neither the platform nor the user (of the service) occupies an institutional position of power that enables them to violate the rights of the worker in question.

Admittedly, therefore, boundary questions - of the kinds presented by the employment/independent contractor binary - will remain. The nature of the boundary questions will, however, alter: instead of focusing upon the traditional indicia distinguishing between employers and employees, and the contractual configuration - which have hamstrung courts in their examination of platform work - focus will shift to institutional differences of power and the precise manner in which the ‘app’ is put to use by the platform.^[573]

An added benefit of this approach is that, in certain instances, the institu­tional relationship might apply to some rights, but not to others. For example, a skilled freelancer may be able to set their own wages (thus obviating the need for a horizontal constitutional right to a minimum wage), but the location of the workplace for that service - say, the service-user’s home - might still require the enforcement of the right to safe working conditions. The institu­tional approach is thus flexible enough to account for situations where, within the labour market, private parties might occupy different locations, depending upon the right at issue.

Now, an objection may be raised here that this approach is too flexible, and presents rule-of-law problems by undermining certainty in adjudication. Indeed, as ACL Davies points out, the rule-of-law-based objection to horizontal appli­cation, which was discussed in chapter five, carries particular salience in labour law, as ‘workers’ rights are likely to be violated by employers, most of whom are private individuals or firms’^[574] (and therefore have rights of their own, as well as an entitlement to a clear understanding of their obligations).

In response to this, first, one may note that in many cases, there will be an existing labour law framework (which may or may not incorporate the insights of the institutional approach) that will set out the rights and obligations of the parties to a platform contract.

For example, in many jurisdictions, there now exists an ‘intermediate cate­gory’ of ‘workers’, which is defined more broadly than ‘employees’, and which is granted a reduced catalogue of labour rights. As I shall show with a few illus­trations in the next section, the institutional approach can be useful in order to both justify the inclusion of platform workers within the intermediate category and critique the situations in which this category is incomplete or insufficient. By contrast, in the context of labour law, constitutional litigation of the kind discussed in chapter five, where legislative vacuum required the declaration of an immediately enforceable, horizontal constitutional right upon private parties, is more likely to be an exception (albeit, a necessary exception) rather than the rule.

Secondly, it is important to reiterate that the purpose of the institu­tional approach is to assist in determining what rights workers have against employers (and, specific to this chapter, against platforms). This is the role that is currently played by the employer/employee binary, and the traditional indicia to distinguish between the two. The historical record shows that the employee/contractor framework has been, and continues to be, the subject of extensive litigation. As cases around the world indicate, this is especially true of platform work, where the employee status of platform workers is consistently brought to court, often with differing verdicts being handed out. Indeed, as Davies notes in the context of the UK,

the tests applied by the courts to decide who is a typical employee and who is not are so ambiguous that researchers may not be able to classify people with any certainty even when presented with the details of their employment relationship.^[576]

In this context, I would argue that the application of the institutional approach, in the manner that has been spelt out so far in this chapter, would not raise significantly fresh challenges for adjudicatory bodies, or present a radical

The Institutional Model as Applied to Platform Work 157 departure from the kinds of issues they are already dealing with, within this domain. Indeed, in the context of platform work, the application of the insti­tutional approach, with its focus on the ‘app’, may bring greater clarity than presently exists.

It can be seen, therefore, that the institutional approach provides a strong theoretical foundation for the indicative factors set out above. These factors are based upon an understanding of what may constitute a genuinely horizontal relationship between two parties. Or, in terms of the institutional approach, they help identify situations where the relevant institutional locations of the two parties within the labour market do not place one in a position of power over the other. In the circumstances outlined above, the ‘app’ plays a relatively marginal role in determining entry into, conditions of and exit from the work relationship. Instead, the ability of the worker (because of factors of skill, specialisation and ownership of business) places them in a position where, arguably, the contract no longer masks an institutional imbalance of power, but is a genuine reflection of a substantively equal relationship.

I will now sum up the argument. The institutional approach provides us with a set of conceptual arguments to justify the application of horizontally appli­cable, constitutionally guaranteed labour rights. It does so by identifying the labour market as a relevant institution. Further, it operates upon the premise that within this institution there exists a difference in power between employers and workers that flows from the former’s ownership and control over capital. This argument then applies with equal force to platform work, along with the following advantages.

First, it provides an overarching constitutional standard that can engage productively with existing labour law frameworks (especially in cases where individuals have been excluded from protection under existing statutory labour law regimes); and, in the absence of an available legislative framework, it makes the case for direct horizontal application of certain labour rights within the plat­form economy.

Secondly, the institutional approach critically interrogates the employer/ independent contractor binary that has structured traditional labour law, and as proven to be specifically problematic in its extension to platform work. This is a suggestion that has been made by other scholars. In a similar formulation, Alan Supiot (along with many others) suggests replacing the employer/contractor binary with concepts of ‘dependent’ and ‘independent’ work, focusing specifi­cally on ‘digital tools that allow owners to control the work of others without giving them orders’,^[577] something that I have discussed above. The institutional

approach does not use the same vocabulary as Supiot, but the overlaps between concepts of institutional power and dependency - both of which, in the context of platform work, place the role of the ‘app’ at the centre - are evident.

And thirdly, because of its focus on how power relations are structured, the institutional approach is able to capture the core features of the platform econ­omy (such as the central role of the ‘app’), and can also dispense with features that are irrelevant (such as the ability to switch between multiple platforms, and ownership of equipment). It is thus in a position to explain when labour rights should apply - as constitutional, horizontal rights - to the relationships between workers and platforms and, likewise, when, and to what extent, they should not.

VI.