Towards privatized social and employment protections in the platform economy? Evidence from the UK courier sector

Platform capitalism has facilitated the widespread replacement of employment contracts with contracts for services. These offer significantly fewer social and employment protections for the independent contractors engaged. What does this mean for the future of national social and employment protection (SEPs) systems? We show how the question of platform workers ’ employment status – and therefore access to SEPs – remain unre-solved under UK law. Drawing on socio-legal theory, we demonstrate why digital labor platforms represent a challenge to existing modes of employment law and labor market regulation. In the absence of immediate legal ‘fixes ’ , some unions and firms are innovating new ‘privatized social protection systems ’ . A ‘Self-Employed Plus ’ (SE + ) agreement in the UK parcel courier sector developed between Hermes, a UK-based courier service, and the GMB union represents an important example of such attempts being made to bridge the current regulatory void. We critically analyze the agreement and draw lessons for platform governance theory. We demonstrate that privatized SE + provisions potentially offer significant benefits for platforms by reducing regulatory oversight, boosting productivity, and enhancing managerial control over platform complementors. At the same time, while they risk undermining national SEP systems and degrading worker protections, they also offer a window of opportunity for trade unions to gain a foothold in the platform economy.


Introduction
This article examines three trends increasingly prevalent across advanced economies' labor markets: (i) the rise of digital platforms intermediating 'transactions' between independent contractors and customers; (ii) the contingency of the traditional contract of employment under such circumstances, alongside its status as 'gatekeeper' to social and employment protections (SEPs); and (iii) the rise of privatized SEP systems, provisioned by platforms to self-employed contractors. We develop a theoretical and historical explanation, rooted in a socio-legal account of platform capitalism, for the reasons behind increasing deployment of privatized SEPs. We examine in detail the functioning of one such system in a UK-based parcel courier firm.
Advances in digitalisation, algorithmic applications, and AI's development into a viable general purpose technology (GPT) have witnessed the emergence of large-scale platform firms across the service sector (Trajtenberg, 2019). Disruptive innovation led by platforms' rapidly scalar business models has transformed areas like hospitality (AirBnB), catering (Deliveroo, JustEat), transportation (Uber) and the logistics/ courier industry (Instacart, DoorDash). Algorithms and AI enable such platforms to match capital and labor with negligible transaction costs, and to manage labor with minimal human interfacing (Lobel, 2018). This, in turn, both permits a commodification of service provision, while effectively eliminating the HR functionfacilitating the replacement of direct employment relationships with complex multiparty contractual arrangements for the provision of services broken up into microtasks (Duggan et al., 2020).
A major upshot of the platform-led transformation of service sector labor markets has been to redesignate the employment relationship as a contract for service provision. This in turn serves to systematically exclude platform workers from access to existing social and employment protection systems, which are largely tied to the traditional contract of employment (from minimum income and paid sick and holiday leave, to the right to trade union recognition, rest breaks, and health and safety at work). In this article, we examine the UK's regulatory environment for the platform economy and its labor markets, highlighting the absence of a functional and consistent legal-regulatory framework capable of addressing ambiguities of employment status and thus SEP provision. In this context, we point to emergent forms of platform-driven SEP provision as examples of a more general increase in privatized forms of governance.
Section 2 reviews the literature and provides context on the emergence of platform capitalism. Section 3 elaborates on regulatory approaches to platform capitalism, and examines how the standard employment contract still acts as a 'gateway to social protection' across all major advanced economies (Forde et al., 2017), and a guarantor of rights inside the workplace. Drawing on socio-legal theory, we develop a sympathetic critique of formalist approaches to lawcharacterizing platforms as a novel form of business organization with distinct abilities to appropriate governance capacities from existing regulatory bodies. We demonstrate empirically the regulatory vacuum which has emerged in the UK's platform labor markets, by presenting an up-to-date review of legal and regulatory developments with bearing on the status of platform workers and, consequently, their ability to access SEPs.
This regulatory vacuum has prompted firms and unions to innovate 'privatized' social and employment protection systems. Section 4 outlines the general features and recent examples of such systems which embed SEPs in service provision contracts with the nominally selfemployed. We subsequently (in Section 5) draw upon evidence from key informant interviews in a landmark agreement in the UK gig economy between the private courier firm Hermes and the GMB trade union, alongside a review of recent significant developments in employment law. We consider the merits and drawbacks of such 'privatized' forms of SEP provision. Section 6 concludes, noting that voluntary SEPs developed by firms and their social partners present opportunities for platforms to deflect regulatory intervention, increase productivity, and enhance managerial control without compromising the self-employed status of their workforces. We further observe that the patchwork and unilateral nature of such agreements present shortcomings versus statutory systems, of which policymakers should be aware, while they also offer distinct windows of opportunity for trade unions.

Background and literature review: platform capitalism and labor markets
In this section we elaborate on the labor market trends associated with an emergent platform capitalism. Our principal empirical focus is on the UK economy, with an eye to broader OECD trends. 1 First we identify key characteristics of platforms and platform labor, and situate our study within the growing academic literature on platform labor markets. We then present evidence of platform work's increasing prevalence at the bottom tiers of service sector labor markets, which are themselves expanding in their relative significance. Finally, we observe the impact of platform capitalism on social and employment protections systems.
What are platforms and what is platform capitalism? Evans and Gawer (2016) provide a working definition of platforms as assemblages of digital and physical infrastructures, controlled by a firm, which intermediate transactions between discrete user groups (a 'multi-sided market'). By capitalising on network and scale effects to 'capture, transmit and monetize data' generated by such transactions, the platform can further expand. This virtuous dynamic process, if successful, cements the firm at the center of a platform 'ecosystem', the architecture and governance of which permit the platform to further identify and tap additional sources of value from its participants by defining the rules of participation (see also Gawer 2014, Bearson et al. 2019. The utilization of big data analytics, machine learning, and cloud computing is particularly important both in terms of facilitating users' interactions on a large scale, and in modeling efficiencies. Platforms' business success is defined not so much through achieving consistent productivity gains as market capture, minimization of costs, and maximization of transactions. Achieving far-reaching control over complementor activities (as in the Taylorist system of scientific management) becomes less significant, for this reason, than facilitating an ever-greater quantity of such interactions flowing through platform ecosystems. What control that does is exist is exerted by algorithmic confines, rather than personalized managerial relationships (Duggan et al., 2020). The 'modularity' of business/ labor functions designed and co-ordinated by platforms provides complementors with (varying levels of) 'autonomy in how they design, price, and operate their respective modules, as long as they interconnect with others in agreed and predefined ways' (Jacobides et al., 2018, Shipilov andGawer, 2020).
Modularity, however, also renders complementors fungibleinsofar as specific functions are not tightly embodied in particular individuals or firms, but can be carried out by an interchangeable pool of complementors (see Vallas 2019). Algorithmically predefining the scope of platform complementors' tasks and roles endows platforms with the role of 'market-makers', since the platform retains ultimate control over pricing structures and the rules of participation. As such, two key legal tests of employment statusnamely control and personal serviceare evaded by the platform work model.
Following Srnicek (2017), Vallas (2019), Langley and Leyshon (2017) amongst others, we interpret this interconnected set of trends as heralding the emergence of 'platform capitalism': whereby a significant transformation of the structure of (labor) marketsin which platforms act increasingly as centers of power and governance (see 3.3 below)gives rise to regulatory challenges which in turn drive institutional transformations. Platforms are beginning to transform labor markets, as they signify increasingly widespread evasion of existing social compacts via regulatory arbitrage (Drahokoupil and Vandaele, 2021).

Labor markets and labor process
A definite uptick in what the International Labor Organization (2016) refers to as 'non-standard employment' has become increasingly prevalent across OECD economies since 2008: fixed-term and/ or zero-hours contracts, hiring through temporary worker agencies, and 'false'/ 'bogus' self-employment. Some remain sceptical of platforms' impact on the growingly precarious nature of work (e.g., Williams and Horodnic, 2018). Individual platforms can be shown to be expanding very rapidly: the number of active Uber drivers in the U.S., for example, increased by approximately 300% during 2015 alone (Hall and Krueger, 2018).
Identifying the general significance of platform work is not so easy, however. One major representative UK survey conducted recently records a rapid doubling in the size the UK's working population involved in platform work on at least a weekly basisfrom 4.7 to 9.6% between 2016 and 2019 (Huws et al., 2019). Between one-quarter and one-third of these regular platform economy participants obtained more than half of their income from said work (and for 9.4%, it represented their only source of income). Extrapolating from this sample, we can estimate full-time platform workers to represent something in the region of 3% of the UK labor forcea small but not insignificant minority amongst a considerably larger pool of part-time platform workers, and a substantial increase over a short period.
The rise of platform-based labor is associated with increasingly precarious 'gig economy' working conditions (Donovan et al., 2016, Schor et al., 2020. The upshot of the 'fungibility' of platform complementors (workers) finds a parallel in the kinds of labor markets created by platforms. These markets systematically diminish working 1 The UK serves as a useful 'least-likely' case for the emergence of independent contractors providing labor services through platforms, given that Anglo-American capitalism is associated with the weakest forms of social and employment protections for those enjoying contracts of employment (implying a diminished benefit for designating workers as self-employed, compared with the potential risk of employee miscategorisation lawsuits).
conditions by radically reducing barriers to labor market entry (Shapiro, 2020). This accords with other research pointing to intermingling between platform and non-platform labor markets, with a minority of workers serving platforms full-time and a larger majority using platform work to top-up declining incomes from employed work (Drahokoupil and Piasna, 2017). By engaging in regulatory arbitrage, they further diminish the competitiveness of firms utilizing formal employment relationships operating in the same sectors, opening the field to platform dominance across a range of location-based services (like ridehailing and food delivery). Moreover, labor platforms typically possess tight centralization and vertical forms of organization, exerting tight control over how workers interact with other platform complementors and consequently the labor process itself (Saadatmand et al., 2019).

Social and employment protection systems
The largely self-employed categorization of the platform workforce limits its access to both social insurance schemes outside of the workplace, and to employment protections at work. The mounting problem of insufficient access to social and employment protections (SEPs) has been recently recognised by the OECD (2018) and the European Commission (2018). As a result increasing efforts have recently been made to guarantee a minimum of social protection to the self-employed (see Avlijas 2019 for a Europe-wide comparison).
Most OECD economies now provide minimal sick pay for employees and the self-employed alike. But for parental leave, pensions, and unemployment benefits, coverage is far more sporadic (if not non-existent) for independent contractors (Spasova et al., 2017). The same is true for employment protections (health and safety protection, hourly rates, working time restrictions, and other workplace-based regulations). Despite important national differences, the platform (non-) employment model is surprisingly similar across national jurisdictions. 2 This is in good part due to the fact that most advanced economies offer substantial tax breaks for self-employed: savings for 'employers' are estimated at 20-30% (Scheiber, 2018). Because of the low rate of tax contributions, states are, understandably, reluctant to grant the self-employed access to insurance schemes funded by employees and employers (Black, 2020).
Despite attempts at change listed above, then, this highlights the continued prevalence of the employment contract as a 'gateway' policing access to social and employment protections (Forde et al., 2017). Most advanced economies continue to operate binary systems of employment law, categorizing workers as either employed or self-employed. However, some have developed intermediary 'worker' or 'dependent contractor' categories which grant limited access to statutory SEPs to the nominally self-employed (which De Stefano and Countouris, 2019 define as 'modified binary systems'). These hybrid categories offer some, but not all, of the benefits of employee status and are targeted at those with less fixed relationships to a given employer. In the UK context, for instance, 'limb (b) workers' (named according to a stipulation of the Employment Protection Act of 1996) are formally self-employedwhile receiving many of the same SEPs as the employed (Adams et al., 2018; see also Table 2 below). As we discuss below, many recent legal cases have turned on whether platform workers are fully self-employed or deserve to be categorised as (limb [b]) 'workers'. In the UK, a binary tax system (distinguishing simply between the self-employed and employees) is distinct from laws relating to employment status, meaning that cases examining the tax responsibilities of those with (intermediary) worker status must be determined independently (Adams et al., 2018). The loss to the UK tax base from the use of self-employed labor is estimated at £5.6bn in 2019-20 (Freedman, 2020). Wood et al. (2019) deem the use of self-employed labor by platforms as the disembedding of labor from the protective systems designed to secure and facilitate its 'social reproduction'; i.e., the ability of workers to continuously return to work, alongside their ability to successfully raise the next generation of workers. A corollary of successful social reproduction is found in firms' ability to retain labor over long periods (Choonara, 2019). Since all work is subject to some degree of on-the-job learning and skill development even in the platform economy, labor retention generally enhances productivity (Meijerink and Keegan, 2019). The employment contract guarantees the reproduction-retention function through establishing minimum statutory SEP provision (by both government and employer) as a condition of employment. But where self-employed labor is used, firms can adjudicate for themselves the optimum SEP provision required to retain workers. In the absence of suitable SEPs, then, private SEP provision can then be conceived as fulfilling a labor reproduction-retention function for digital labor platforms.

Regulatory frameworks for platform capitalism
This section examines in more detail attempts to regulate platform capitalism. We offer a sympathetic critique of this literature as insufficiently sensitive to the broader transformation in business systems and political economy represented by the emergence of the platform model. We then mobilize recent advances in socio-legal theory to outline an expansive understanding of 'platform governance' as reaching beyond the scope of state regulatory frameworks. In a second section, we review the 'regulatory vacuum' presented by recent legal developments addressing in the UK platform economyutilizing our socio-legal framework to demonstrate how the UK courts have struggled with the classification of platform workers.

Legalistic approaches and a socio-legal alternative
Analyses of the platform economy's regulatory environment is characterized by detailed attention to the contours of platform capitalism's engagement with divergent national legal systems (e.g., Lobel, 2017, Prassl, 2018, De Stefano and Aloisi, 2019, De Stefano and Aloisi, 2018, Donovan et al., 2016. Sustained scholarly attention has rested upon untangling the specific means and devices by which platforms bypass employment laws and regulations in their engagement of independent contractors. These have been variously referred to as forms of 'regulatory entrepreneurship' (Pollman and Barry, 2017); 'legal hacking' (Alexander and Tippett, 2017) and 'regulatory arbitrage' (Pollman, 2018).
Debate has focused on whether bifurcations in labor markets are best ameliorated through a generalized reduction in employment protections in order to facilitate inclusion, or by levelling up employment protections in order to protect the quality of employment (see Rubery and Piasna, 2016). Most scholarship takes the latter approach, and various (and variously convincing) legal 'fixes' along these have been advanced Kountouris, 2011, Stewart andStanford, 2017). These range from placing the definitional onus for legal tests on the role of the employer, rather than the worker (Prassl, 2015), to clarifying the boundaries of competition law to exclude platform workers (Lianos et al., 2019), and from scrapping intermediary worker/ dependent contractor statuses in favor of a firm boundary system with a stronger test of employee (Ewing et al., 2019) to 'reweighting' legal tests of employment status to favor workers (Deakin, 2020).
Why, then, have jurisprudential and regulatory solutions to labor market regulation not been taken up, even as the scale of selfemployment in the platform economy continues to grow? Restricting the understanding of labor market outcomes to the results of the (non-) deployment of statutory instruments or faulty interpretations of legal code or precedentcrucially important though these factors may berisks obscuring the broader sociological trends which co-evolve with and structure developments in legal interpretations, applications, and corporate governance practices (Mather, 2011).
We propose here to apply a socio-legal approach to resolve this question. A socio-legal approach to law and regulation seek to situate the latter two forms within a broader sociological framework (Perry--Kessaris, 2015, Creutzfeldt et al., 2019). 3 It takes as its starting point that the state (alongside its systems of laws and regulations) forms only one normative source of authority within any given socioeconomic system. The analytical focus shifts from how the state regulates society, to how law and society relate to one another in distinctive historical periods and discrete political territories (Edelman and Stryker, 2010, Swedberg, 2003, Teubner, 1993 It further emphasizes the intrinsically reactive nature of law to already-existing socioeconomic developments (as Pierre, 1986 writes, 'the shaping of practices through juridical formalization can succeed only to the extent that legal organization gives explicit form to a tendency already immanent within those practices'). And finally, socio-legal frameworks can broaden our conception of governance to include how multiple and competing normative rules (beyond the law) structure norms and shape behavior (Boltanski andThévenot, 2006, O'Reilly et al., 2014). Adopting this approach clarifies two important points given the context of this study. First is the historically contingent nature of the standard employment contract and its relationship to SEP provisions. Second is the potential to better conceive platforms' growing role as sources of regulation, governance, and law-making.

Employment contract to contract for services
The standard contract of employment is a historically and geographically specific legal form. It developed to enhance managerial control in factory settings where the close supervision of a foreman was necessary to achieve productivity enhancements, before being transplanted to white-collar labor markets as a means of facilitating career progression (Mummé, 2015). These conditions no longer hold under platform capitalism, which is content to rely upon algorithmically managed, deskilled labor earning a piece-rate (Fudge, 2017). Neither does the standard employment contract approximate anything like a norm on a global scale (and cannot then in this sense be understood as a 'standard' form at all: see Munck et al. 2020, Benanav 2019.
The nexus linking the provision of social and employment protections with the standard contract of employment is equally historically contingent. Minimum statutory SEP provisions, embedded in employment contracts, formed a key measure of the neoliberal governance paradigm in industrial relations (Peck and Tickell, 2002). The uprooting of sectoral collective bargaining agreements between trade unions and panels of employers (almost 90% of workers were covered by such agreements in 1980, down to just 27% in 2020) was ameliorated by the introduction of a new individualized institutional framework intended to guarantee the social-reproductive capacities of labor. The construction of a link between employment contracts and SEPs thus tacitly acknowledge the withdrawal of organized labor from the industrial relations framework. It tied minimal in-work and social protections to individual contracts of employment which could, if necessary, be pursued through an individualized employment tribunal system (or preferably resolved through internal workplace procedures) (Pollert, 2005, O'Sullivan et al., 2015. The distinguishing feature of employment law for Kahn-Freund lay both in its legal recognition of unequal stature of the parties involved, and in providing a 'countervailing force to counteract the inequality of bargaining power which is inherent and must be inherent in the employment relationship' (quoted in Davies and Freedland, 1983). Its distinctiveness rests in its incorporation of redistributive principles and attentiveness to power imbalances. Platform capitalism has developed precisely as a means to exploit the interstices of existing legal frameworks to bypass this inconvenience (what Lobel, 2016 terms 'definitional disruption'). Recasting the long-run employment relationship as a series of contracts for the provision for individual services, platforms have been able and willing to replace employment contracts with service contractsan area of law where courts are far more reluctant to adjudicate according to redistributive principles and a formalist approach to law predominates (Zumbansen, 2006, Zumbansen, 2007. This has facilitated platforms' orchestration of what Behling and Harvey (2015) refer to as 'instituted competition [our italics] between different qualities and types of labor' -full-time, semi-skilled and legally-recognized employee labor, versus (often) part-time, deskilled independent contractors without access to SEPs. This dynamic, alongside the historical contingency of employment law and the employment contract-SEP nexus, together highlight the possibility of alternative modes of orchestrating and institutionalizing the utilization and reproduction of labor.

The platform as an institution of governance
The second benefit of a socio-legal approach is its emphasis on how governance takes non-legal/ statutory forms. Corporations have long worked to supplant public law with private forms of governance (Ciepley, 2013). As Shaffer (2009) notes, 'by usurping the role of external legal processes and supplanting them with internal rules, large organizations can enhance their ability to limit legal change' (see also Edelman and Suchman 1999).
Platforms have been able to substantially increase this capacity (Cusumano et al., 2021), precisely because their disavowal of direct control over agents and ownership of assets poses major problems for regulatory systemswhich rely upon the apportionment of agency to specific social actors. Helberger et al. (2018) draw on Thompson (1980) to refer to this as the 'many hands problem': since platforms are not directly responsible for their users' conduct, liability is dispersed across multiple agents. Contracts, furthermore, cover specific acts of service and not long-run working relationships. These may be circular arguments, but they have proved effective in the eyes of existing law (see section 4, below).
As such, the hollowing out of employment law and the rise of privatized and corporate-defined standards and codes of conduct should not be understood merely as forms of liberalization. As Cohen (2019) argues, what is notable about the platform economy (compared even with earlier forms of privatized governance) 'is the degree to which it has taken on the mantle not of deregulation but of managerial reregulation to prevent different institutional configurations from emerging'. Platforms have increasingly become 'policy hubs in their own right' (Cohen, 2020, Gorwa, 2019. This is at least partially conducted with the consent of governments, which have little desire or capacity to regulate the activities of platforms (although this is beginning to shift). While the role of the state may be minimized, then, platform governance is not unidirectional and can incorporate inputs from a diverse range of social actors: user groups, civil society organizations, and industry bodies (Gillespie, 2017).
Workplace partnerships have not so far played much of a role in platform governance, however. Platforms have led unions to experiment with new organizational models (Moore and Joyce, 2020), and to the emergence of new and radical 'independent' trade unions. The latter, such as the Independent Workers of Great Britain (IWGB) and the United Voices of the World (UVW) in the UK, have specifically targeted firms like Uber and Deliveroo amongst other platform firms using self-employed labor (Gall, 2020). Unions operating in these sectors confront the problem of transient, unskilled and atomized workforces. In the minority of cases where unions have established footholds amongst platform workforces, they have done so through militant and aggressive public relations campaigns, and industrial relations have become highly fractious (Però, 2019). Litigating for employee or limb (b) worker status has also become a major vector of trade union activism within the UK platform economy (Kirk, 2020, Leighton, 2016 not least because of the potential threat posed by competition authorities by attempts to unionize the self-employed (Doherty and Franca, 2020). 4 In 2018, the Danish platform Hilfr.dk signed a collective agreement with the 3F trade union, and in the same year food delivery riders in Bologna developed a charter of employment-style rights for those working within the city (Aloisi, 2019). But such examples of platform governance being opened to trades unions input remain few and far between.

The UK's regulatory ambiguity
The employment status of platform workers in the UK is ambiguous and has not been definitively established, as recognized by the Taylor review into modern working practices (Taylor et al., 2017). UK statute provides a circular definition of employee under the Employment Rights Act of 1996 (as a person working under a contract of employment), and so criteria for employee (vs self-employed) status has developed by a series of common law tests (see Freedland and Dhorajiwala 2019 for an excellent summary). These comprise, briefly, of control (over the labor process), mutuality of obligation, and personal service (i.e., whether the worker has a right to use a substitute). The tests are the same for both employee and limb (b) worker status, though the fulfilment threshold is (theoretically) lower for the latter.
Unions and individuals have pursued a range of legal claims regarding platform workers' status under these tests (see Deakin et al., 2015 for an earlier example of a litigation strategy). Below, we highlight seven significant pieces of UK case law relating to the independent contractor status of the 'new self-employed' since June of 2018 in the UK (Table 1), bringing up to date prior research along these lines (De Stefano and Aloisi, 2018). 5 This summary of recent legal developments presents a highly contradictory picture. On the one hand, courts have often been willing to rule that nominally self-employed workers are indeed limb (b) workers or employees (Uber v Aslam; Pimlico Plumbers v Smith). On the other hand, such cases do not appear to have established clear precedent, and thus do not (yet) possess substantive implications beyond the parties involved.
For instance, the Employment Appeal Tribunal referred to the Pimlico Plumbers decision to rule in the Universal Aunts case (appearing to establish the principle that practical difficulties of substituting equated to employee status). But the Pimlico judgement had no bearing on the later IWGB v Deliveroo case, which turned on the formal right of Deliveroo riders to use substitutesdespite similar evidence of the impracticality of doing so (see Atkinson and Dhorajiwala 2019). By contrast, the CJEU judgement in B v Yodel asserted tests for self-employmentproblematic under UK law given that there is no default assumption of employment status. Addison Lee similarly noted the Pimlico case but turned on the question of mutual obligation. While the Uber v Aslam case came to an apparently strong judgement, its broader ramifications will only be seen gradually in the myriad cases likely to be brought to tribunal over coming years. The inability to establish a clear legal-regulatory framework has entailed a de facto acceptance of platforms' practice of contracting with the self-employed. In the context of uncertainty over employment status, however, platforms and their workers continue to possess a (partially overlapping) mutual interest in the provision of SEPs. For workers, the benefits of SEPs are plain. Firms, must, on one hand, retain labor (even where costs of replacement are low, they are never zero) ensure its social and maximize HRM potentials for labor control. On the other hand, competitive pressures in the sector disincentivises labor cost increases associated with increased forms of social protection. The necessity to provide some social and employment protections to fulfil this retentionreproduction requirement has led many platforms to begin providing what we term privatized social and employment protectionsembedded in service provision contractsfor the self-employed independent contractors whose work they intermediate as illustrated by the Hermes case in the UK.

Hermes and Self-Employment Plus
As noted in section 3, platforms have been identified as a new organizational form of business, representing increasingly powerful centers of private governance and standard-setting. Private social and employment protection provision is best understood as one emerging element of platform governance, with the effect of displacing formallegal requirements. For example, Uber now offers AXA insurance provision for drivers in the UK to provide £1000 in payments for parental leave, 15 days sick pay and compensation cover for industrial injuries and jury service (AXA XL Insurance, 2020). The 2019 French 'Law on Mobility Orientations' outlines a process (in development) by which platforms must negotiate social responsibility charter outlining obligations to their contractors (Danesi and Li, 2020). The coronavirus pandemic of 2020 accentuated such trends. In the UK, DPD and Deliveroo put aside funding to pay for sick leave for couriers needing to quarantine, while in the US both Lyft and Uber provide limited paid sick leave for those affected by the coronavirus (Booth, 2020). Instacart also recently agreed to payouts in lieu of providing health and sick leave benefits in San Francisco in an out of court settlement with the city's Labor Standards Enforcement (Said, 2020).
To better understand the operation and assess the efficacy of such privatized SEP systems, this section introduces materials from a case study involving parcel courier Hermes UK and the GMB trade union conducted during spring/ summer 2020. We examine the attempt by the GMB to litigate on behalf of courier drivers for working with Hermes for 'limb (b)' worker status under UK employment law. First, we introduce the UK parcel courier sector as an example of emergent platform business model at work, before introducing the facts of the case obtained through original interview research with participants, document analysis and third-party data. We conducted key informant interviews with firm representatives, trade unionists, and practitioners/ experts familiar with the case. Materials related to the new contract signed by couriers, promotional materials, and excerpts of the agreement were also examined. We argue three notable characteristics of the agreement have emerged: the rise of 'privatized' platform governance; the emergence of a 'platform partnership'; and the extended reach of platform managerialism.

The UK courier sector and Hermes as a platform company
The parcel courier sector in the UK has experienced a boom since the advent of business-to-consumer e-commerce, with parcel volumes totaling 3.5 billion generating revenues of over £10bn in 2019 -making it one of the most mature global markets for parcel delivery on a per capita basis (Ofcom, 2019). Returns are slim (averaging around 5%) and competition ferocious. Two big private players established in the late 2000s -Hermes and Yodelcompete with more recently-established Amazon Logistics and the former state monopoly the Royal Mail that continues to handle just under half of all parcel volumes in the UK. Collectively these firms account for three quarters of all parcel volume in the UK, while DPD and UPS dominate the market for the remaining (mostly higher-value) freight. Hermes delivered over 400 million parcels in 2019. Its UK turnover was £860m in the financial year 2019/20, and the firm posted a pre-tax profit of £46.2m (a return of 5.4%, up from 4.8% in 2018/9). 6 Despite protestations to the contrary (DLPA 2017), courier companies, including Hermes, increasingly demonstrate characteristics of the platform business model. Hermes' UK operation can be subdivided into two parts: 'back-end' and 'final-mile'. Hermes' asset-heavy back-end can be understood as the 'platform hardware' (Constantinides et al., 2018). This includes a fleet of around 10,000 vans and 400 heavy goods vehicles (HGVs), alongside a network of parcel-sorting hubs, delivery depots and offices: fixed assets with a total value of £117.8m in February 2020. A new 'superhub' -capable of handling 1.3 million parcels dailyis planned to open in Barnsley in the North of England in 2022, at a cost of over £60m. Hermes directly employs around 3000 staff, a mixture of white-collar staff, HGV and van drivers, and hub and depot parcel sorters. Field managers oversee the day-to-day business of hiring couriers and facilitation delivery operations for multiple local rounds; they are directly employed by and accountable to Hermes.
But as with other courier firms, beyond this asset-heavy core lie characteristics strongly identifiable with the platform business model. 7 Hermes' final-mile operations are principally carried out by around 20,000 self-employed couriers. This 'external' labor force greatly etc. is the major role of speculative venture capital in financing the latter ('classic') platform firmsa distinction which produces no significant functional difference in terms of corporate governance, the labor process, or the employment status of couriers.
outnumbers Hermes' circa 3000 employees. 8 To cope with the surge in demand from the Covid-19, Hermes expanded by hiring 1,500 additional employees, alongside engaging 9000 extra self-employed couriers through 2020. Lifestyle couriers collect parcels from outsourced subdepots, local sorting and collection sites leased and operated by selfemployed contractors (sub-depot controllers). Couriers own and are responsible for their own vehicles. Courier work is broken up into geographical rounds by a software tool. Rounds typically cover a single UK postal code.
Most couriers sign open-ended service provision contracts assuming responsibility for delivering parcels on a given round and are designated as 'roundholders'. Couriers hold on average one and a half rounds each, while some hold up to ten (generating incomes of over £100,000) and subcontracting some or all of these to other (nonroundholder) couriers. Couriers collect parcels from a Hermes sub-depot each morning at a set time and scan them using a smartphone app (previously a handheld terminal). Hermes is in the process of minimizing contact between field managers and couriers, in favor of automated app-based problem solving (e. g., logging a missing parcel or the use of a substitute). Maximizing the potential for control over parcel flow, it is using geolocation data to allow customers to geofence safe delivery spots where couriers must enter and photograph the parcel in order to record their delivery and receive a receipt. The IT system was fully transferred to an external cloud server in February of 2020, permitting a rapid scaling up of delivery services during busy periods. Further IT development is aimed at improving the capacities for customers to redirect and/ or reschedule deliveries at ever-shorter notice periods. Expanding the number of platform complementors is being pursued through establishing partnerships with existing small convenience stores to handle growing volume (ParcelShops) and the deployment of over 800 Parcel Lockersboth of which greatly enhance consumer-to-consumer (C2C) parcel volumes. This increased business focus on software development and big data analytics is reflected in a rapid rise in the value and relative significance of Hermes' intangible assets: from £18.9m in 2018 to £44.5m in 2020 while the value of its tangible assets remained static during this period (Hermes Parcelnet Ltd 2020) In sum, Hermes' reliance upon a platform model is evident in its: (i) Algorithmic assignation of work (routes and timings are modeled by a VRP mapping algorithm for pay claim purposes) (Leyland and Others v Hermes Parcelnet Ltd, 2018); (ii) Piece-rate payments for micro-tasks (a per-parcel rate, with performance incentives); (iii) Organization of work through digital infrastructure and by algorithmic management; (iv) Intermediation of multiple user/complementor groups (couriers, retailers, consumer-senders and consumer-recipients); (v) Capture and utilization of data generated by network and scale to generate efficiencies; (vi) Minimization of fixed capital ownership and direct employment, relative to external (complementor) assets.
The growth of business-to-consumer (B2C) e-commercewhich sits at the core of Hermes' operationseffectively displaces jobs in logistics delivery sector serving physical retail outlets. Employment in the logistics sector is then effectively replaced by self-employment in the parcel courier sector.

Introducing the self-employment plus contract
Hermes' lifestyle couriers have always been self-employed, and consequently exempt from employment and social protections tied to the contract of employment. After recruiting several members inside Hermes, the GMB trade union embarked upon litigation at an employment tribunal in 2017 aimed at securing limb (b) worker status (an intermediary category permitting paid holidays, sick pay and a minimum hourly wage) for a small number of couriers. 9 The Employment Tribunal found in favor of the litigants in June of 2018, and the GMB noted its intention to raise claims on behalf of more couriers. Hermes, for its part, planned to appeal the ruling. Legal proceedings were dropped in early 2019 when what both parties termed a 'landmark agreement for the gig economy' was announced: a new 'self-employed plus' (SE+) category to which Hermes drivers could opt-in on a voluntary basis. The status has since been made mandatory for new starters and a full conversion process of all couriers is underway, albeit delayed by the onset of the Covid-19 pandemic.
The agreement consists of three main components. First is a union recognition agreement with the GMB, securing its right to formally represent those couriers on an SE+ contract. This agreement consists of a 'Partnership Principles Agreement' (outlining the principles of a cooperative approach between company and union) and a 'Procedure Agreement', which establishes the organisational structures and their functions. These include a Joint National Negotiating Forum (JNNF), a Joint National Consultation Forum (JNCF), and local Regional/ Divisional Consultation Forums (RDCFs).
Briefly, the JNNF negotiates changes to SE+ couriers' pay and holiday scheme. It meets annually and consists of members of Hermes' senior management and nine GMB negotiators. The JNCF consists of the same membership as the JNNF but is scheduled to meet quarterly, to discuss and resolve issues arising from routine business operations that affect couriers, and to resolve lower-tier contract disputes escalated from the RDCF. RDCFs are localized institutions including Hermes Divisional (i.e., regional) and Field Managers and local GMB officials and representatives. These are held monthly and aimed at local dispute resolution and the identification of issues for escalation to higher forums. The agreement further contains stipulations regarding facility time for GMB representatives, union-led health and safety auditing mechanisms, conditions of access rights for the purpose of union recruitment, and a right to trade union representation at RDCF (but not at the national scale) for Lifestyle Couriers (i.e., those not on an SE+ contract).
Second is the minimum hourly wage guarantee. The legal hourly minimum wage for adults over 25 (National Living Wage) is £8.91 as of April 2021. The SE+ contract states that payments 'in respect of the Rates, bonus payments and the SE+ Admin Payments shall be equal to or more than the National Living Wage in each relevant period.' Payments to such effect are automated. The SE+ payments software built upon already-existing pay modeling software. However, prior to the introduction of SE+, and despite the existence of software, such payments were essentially made at the personal discretion of Field Managerswith Hermes' Director of Legal and Public Affairs (2017) noting that 'Hermes Field Managers intervene to make ad hoc payments to ensure that the courier's earnings are not harmed'. Interviews with trade union officials and members corroborated this account, with one claiming that while low-pay was endemic for couriers prior to the advent of SE+, the company had greatly improved since the introduction of the new pay mechanism (Interview U3). Also automated are various 'incentive payments' (e.g., £2 for delivering 90% of parcels within set time windows), for which couriers previously had to pursue claims with Field Managers.
The entitlement. UK employees and limb (b) workers are entitled to 28 days paid annual leave, while the self-employed are not entitled to any paid or guaranteed leave. Hermes Lifestyle Couriers must provide their own cover and thus in practice very rarely take holidays (Interviews U1; U2). As part of the SE+ agreement, Hermes guaranteed SE+ couriers 28 days paid leave per annum. SE+ couriers accrue paid leave for each full day worked (regardless of whether a substitute is used). The system 'works like a bank' (interview H4) insofar as a piece-rate pay reduction is taken in exchange for holiday accrual at a rate of approximately 0.1 days leave per day worked. Hermes developed an online Marketplace system couriers use to book leave (at least four weeks in advance). This also advertises uncovered rounds to other couriers in the area wishing to take on temporary additional work. At present, around 2000 couriers work on the SE+ contract, while the remainder (around 10,000) remain as 'regular' self-employed Lifestyle Couriers. However, as noted above, management intends to move all round holding couriers onto SE+ contracts. Table 2 presents the key differences between UK employment models and the provisions granted SE+ couriers by virtue of the contractual agreement.

Deflecting regulatory challenges
Private provision of social and employment protections through the creation of SE+ status is evidently intended to deflect the growing regulatory challenge to Hermes' platform employment model. Recognition of the trade union and the introduction of the new contract have served to check the immediate legal threat. These steps also demonstrate an intention to "symbolize compliance" with employment law, while modifying its scope and application to better suit business requirements (Shaffer, 2009). Achieving symbolic compliance consists in the complex task of designing an SE+ contract that refute evidence of an employment relationship, while replicating certain statutory SEP provisions in order to repel regulatory encroachments.
This latter effort at symbolic compliance is evident in how two core SEP provisions associated with limb (b) worker statusthe minimum income and paid leave guaranteesare effectively replicated by the SE+ contract. It is also evident in the design of a £1m sick pay fund introduced for couriers shielding, taken ill, or aged over 70 during the Covid-19 pandemic, which paid these groups £20 per day for up to 28 days. Discussing this specific initiative (but making a point more generally applicable to the contractual status), a member of the Hermes legal team noted that it would be 'hugely brave and unlikely of HMRC to criticize, [to] subject to a disadvantage, a company that, again, had thought for themselves and done something themselves… the two legal and the direction of public opinion arguments are why I'm fairly confident that that will be [safe from regulatory challenge]' (interview H2). In this way, SE+ is intended to signal the adequacy of SE+' SEP provisions to regulators.

Averting future litigation
The text of the SE+ contract, by contrast, is carefully designed to avert any potential future litigation over couriers' legal status by denying the existence of an employment relationship. The agreement (alongside Hermes' managerial practices for couriers) was designed in collaboration with a senior QC, with the goal 'of making sure they [couriers] stay self-employed […] for tax purposes [and] employment law purposes' (interview H1). Section 7.1 of the SE+ contract states: 'You are not an employee of Hermes and you accept and agree that you provide the services to Hermes as a self-employed person'.
Averting working time regulations, Section 4.1 specifically states 'The Working Time Regulations 1998 do not apply' (referencing to a key legal definition of limb (b) worker status). Finally, the contract makes clear that couriers enjoy an unfettered right to deploy a substitute ('5.1 You have the unfettered right to identify, engage and pay a substitute to perform any element of the services… under this agreement'). The holiday booking Marketplace and the leave system specifically address an aspect of the Employment Tribunal hearing. The judge noted that placing the onus on the courier to find a substitute amounted to an obligation to conduct personal service; the new system places the onus for finding a replacement during booked leave upon Field Managers.
A Hermes management representative summarized the design process of the SE+ contract as follows: 'Looking at the tests of employed employment status, which are control, mutuality of obligation and substitution, and making sure the design of SE+ still means that we tick [i.e., avoid] at least one, possibly two, of those boxes... So, that's why substitution is, kind of, really key to our model' (Interview H2).

Mutual benefits of bypassing the courts system
Interestingly, given that the union had instigated litigation, both Hermes and GMB representatives acknowledged the mutual benefits of bypassing the courts system in striking the agreement. A Hermes manager expressed the potential positives of an out-of-court agreement, Table 2 SE+ social and employment protection contractual provisions, in comparison with employee, limb (b) worker, and self-employed statutory provisions. does not as yet cover day-to-day illness. ** Health and safety protections do not replicate statutory provisions, but simply introduce a union-led safety auditing procedure following an incident involving a courier.
noting that whatever the outcome of the employment tribunal, the company and union would 'probably end up talking anyway. The official also dismissed the significance of worker status, claiming that a union recognition agreement and an active network of shop stewards was more valuable than formal statutory limb (b) worker status. This was principally due to the leverage given to the union by the negotiating forums and grassroots pressure. This claim was reinforced by the account of two shop stewards, one of whom had been part of the tribunal claim which established their status as worker with Hermes, and another on a who claimed there was no substantive difference in their terms or conditions of work (interviews U3, U4).

Trade union and emergency partnership
The preference of firm and union for an in-house agreement highlights a second theme emerging from the evidence analyzed: the surprisingly rapid development of a strong firm-union partnership. During the height of the UK lockdown in April and May of 2020, as consumers shifted rapidly to online retailers, over 2.5 million parcels were handled during a single daysubstantially surpassing a record set during the previous year's winter peak. The intense business demands imposed by the Covid-19 pandemic rapidly accelerated co-operation under the Partnership Agreement, driving what we identify as a form of 'emergency partnership' between Hermes and the GMB.
Emergency partnership relations manifested in substantially increased regularity of meetings between senior management, union leaders and shop stewards. The Joint National Consultation Forum (JNCF) moved its quarterly meetings to weekly Monday morning conference calls beginning in late March 2020. Here a network of shop stewards reports in real time problems experienced on the ground in sub-depots and by couriers. The speed at which this relationship developed took even senior trade unionists involved by surprise: '[The GMB] is years ahead of where we thought we would be… One of those things has been the complete and utter collaboration in turning Hermes around with its responsibilities to Covid-19' (interview U1).

Increasing legitimation and reputation
The presence of a shop floor union presents management with an opportunity to project an image of corporate responsibility. As one shop steward stated: 'when we talk to the company at the very top, they will not tolerate [poor management]… they have actually… new clients talking to them because they have cleaned up their act' (interview U2). Another trade unionist strongly emphasized the partnership approach being taken by the GMB, recounting a negotiation with management where stated 'you think that as a trade union we want to burn your business down… what we want is a collaborative approach to actually, how can your business still make money, but do it without exploiting people and treating them with respect' (Interview U1).
From the union's perspective, the partnership arrangements have resulted in numerous cases of work being reinstated to couriers who had had rounds removed, and successful pay disputes being resolved at RDCF and the JNNF. Furthermore, several Field Managers and sub-depot controllers have been subjected to disciplinary notices by senior management for their interactions with couriers, as a result of GMBfacilitated complaints procedures (Interviews U1, U3 and U4.) The desire to placate new majority shareholders Advent Capital, a private equity group which acquired 75% of the company in August 2020 from Otto Group, was also highlighted as an incentive for Hermes to avoid having the GMB issue a 'failure to agree' during pay negotiations in the spring of 2021. The offer ultimately negotiated was accepted by 94.5% of voting members in March that year.
Such partnership arrangements are quite unique in the platform economy, which elsewhere is characterized by both highly antagonistic and informalized patterns of industrial relations (Cant, 2019). Nevertheless, clear conflicts of interest remain between union and company. One is over the nature and scope of the agreement. While depicted by senior managers as being overwhelmingly to do with the right to paid leave ('ultimately, it [SE+] just boils down to paid holiday… three or four months of detailed negotiation boils down to the one thing, really' [interview H1]), trade unionists viewed it as a more expansive framework by which to achieved further improved terms, pay increases and formalized rate-setting over a multi-year period (interview U1). One estimated that repayments to couriers secured by the GMB amounted to around £4.2 million over a two year period.
A second potential antagonism lies in the scope of the agreement. Cover (part-time) Couriers and Lifestyle Couriers (those not signed up to SE+ for whatever reason) remain outside SE+ arrangements. A round holder hired one such courier during the peak of the Covid-19 pandemic. He was not aware of the union, complaint/ dispute resolution procedures, or the existence of additional incentive payments beyond the perparcel rate. He left his role after three months as 'work kind of dried up' during the early summer of 2020 (Interview C1). Another courier regularly responsible for a subcontracted round who had been working over a longer time period displayed similar lack of awareness (interview C2). Such experiences imply that the partnership framework could lead to the emergence of a segmented internal labor market for couriers: a core workforce benefits from income security, guaranteed leave and union protection of SE+, while Lifestyle and Cover Couriers possessing diminished pay and conditions are subcontracted during peak periods. Hermes' crowdworking model (and the fungible, superfluous characteristics of this workforce identified by Van Doorn [2017]) are as such not necessarily threatened by the agreement.

Refining platform managerial functions
A general tension of the platform business model is to avoid employment law tests of subordination while maintaining managerial control over labor to uphold service quality. The SE+ agreement enhances Hermes' managerial capacities in two critical ways. First is a heightening of managerial control over lower tiers of the organization managementprincipally Field Managers and Sub-Depot Controllersby virtue of the partnership framework discussed above. Second are increased managerial capacities to control courier work, through the rationalization of scheduling and greater influence over pay. In both areas, the SE+ agreement and partnership framework can be seen to enhance the orchestration capacities of the platform over its complementors.
The empowerment of couriers' union representatives effectively deputises them in service of Hermes' corporate restructuring processes. A GMB official explained this dynamic referencing physical improvements made to infrastructure and hygiene practices at sub-depots following union reporting: 'we've cleaned up subdepots, we've found out weaknesses… … we persuaded them to send out instructions to sub-depots' (interview U1). Another union representative reported two Field Managers for attempting to scupper courier transition to the SE+ contract, each of whom subsequently received disciplinary notices from senior management (interview U2). He observed that senior management had granted unlimited access to all sub-depots nationally without restriction or notice to union reps and officials for recruitment purposes, given that union recruitment went hand in hand with the switching of couriers to SE+ contract. And with regard to delivering complaints regarding missed payments for couriers, union intervention appeared to frequently highlight problems with IT systems of which the company was not previously aware.
This complementarity of interests was also highlighted, albeit less explicitly, by a member of Hermes' management team, who noted it was sometimes difficult to 'reassure all of our field managers' about the legitimate role of the union and SE+ status, and that 'we don't just have to win over the couriers... You've got to win over your own people as well, and any cynicism that they've got' (Interview U3). For this reason, a video promoting the new contractual status was produced in order to facilitate senior management's direct communication with couriers (unmediated by lower managerial levels), and a copy of the contract mailed out to all round holders with the option of converting their status.
Despite the 'symbolic compliance' observed above, the SE+ agreement represents a de facto acceptance (absent of a formal employment relationship) of mutual obligation/ relations of dependency between couriers and Hermes. A senior manager effectively accepted this by contrasting SE+ couriers with part-time Cover Couriersthe latter portrayed as a group who have 'got their own thing that they are doing. They are not dependent on Hermes for their day-to-day income' (implying that the former are indeed economically dependent upon Hermes; Interview H1).
Platform-based human resource management (HRM) practices (as identified by Meijerink and Keegan 2019) are evidently enhanced by the SE+ agreement, including greater control over scheduling power and rate-setting. As one shop steward noted, 'on a practical basis what it [SE+] can do… for [Hermes] is they know, now, when people are going to book holidays' (interview U2), while another union representative noted the on-going work to rationalize pay-bands into inner urban, outer urban, and rural bandings, given the substantial differences in delivery times across geographical contexts (interview U1).
The minimum income and paid leave guarantees are significant in this regard because of widespread evidence of couriers earning below the National Living Wage and rarely being able to arrange suitable cover during holiday periods (Interviews U3 & U4; Leyland & Others v Hermes Parcelnet Ltd, 2018). As noted above in Section 2 with reference to the reproduction-retention HRM function, stronger national SEP regimes broadly correlate with enhanced economic performance (Dunford, 1996, Deakin, 2016. Hermes relies upon couriers' tacit knowledge and learning-by-doing for performance (Interview U2). This incentivises the firm to address relatively high rates of courier turnover, to which SEP provisions can contribute. This productivity-enhancing potential of SEPs was acknowledged by a manager, who noted that 'anecdotally, you know, we hear that couriers feel more engaged. So we have less churn'.
The manager also signaled the existence of a data analysis project aimed at making a business case to the board of directors for maintaining (and possibly expanding) the scope of such provision (Interview H5). The increased profile of the trade union within the firm notwithstanding, there is little evidence of any substantial increase in pay or reduction in work intensity for couriers. This is in part due to the patchwork nature of SE+ -it has not yet impacted per-parcel/ package rates or attempted to regulate working hours, and sick pay does not form part of the original agreement but an exception that has only emerged during the Covid-19 pandemic. Indeed, as a Hermes senior manager noted, SE+' holiday pay 'works like a bank' (interview H2) rather than comprising any additional outlay for Hermes. The accumulation of accrued leave, in exchange for a lower per-parcel rate, incentivises couriers to remain with Hermes.
In sum, SE+ can be identified as an emergent form of platform governance. It fulfils three strategic functions. First, it averts the legalregulatory challenge by extending a tranche of SEPswhich resemble legal/ statutory normswithin the service provision contracts offered to drivers (thus replacing public with private, platform-based, regulation). Second, it fosters the emergence of an emergency partnership with the GMB trade union and its members, pre-empting the potentially disruptive industrial relations apparent elsewhere in the platform economy. Thirdly, in so doing, it extends the managerial control of the platform both (1) over couriers (through rationalizing holiday entitlements, cover arrangements, and rates of pay) and (2) over lower tiers of management (Sub-Depot Controllers and Field Managers), who are subject to heightened scrutiny from couriers and shop stewards' informational relays to senior management. All of this additional control is, crucially, achieved without compromising the self-employed categorization of Hermes' couriers.

Conclusion: towards privatized protection?
The governance functions appropriated and exerted by platforms, and the regulatory strategies states and regulators should employ in response, are a growing area of research (e.g., Pistor, 2020, Jacobides and Lianos, 2021, Duch-Brown et al., 2022. This article has examined the relationship between platform capitalism's emergent legal-regulatory regime and the rise of privatized social and employment protections systems, provisioned through private contracts for services. Legal scholars have principally focused on the mechanisms by which platforms 'hack' existing employment law to position laborers as self-employed independent contractors without access to social or employment protections. To this end, they have recommended legal fixes aimed at challenging this designation and upholding the employment contract-SEP nexus in the platform economy. Courts and governments have, however, not taken up the challenge and a regulatory impasse has ensued in which the status of platform economy workers remains unclear. Our theoretical framework highlights more fundamental forces at work, by connecting insights from socio-legal studies with the literature on platform economies, governance, and labor markets. Foremost amongst the trends we identify is the reorganization of the principal form of the capitalist business organization, from traditional corporation to platform (Davis, 2016). The platform moves away from the ownership of assets and control over employees characteristic of the multidivisional corporation. Instead, platforms aim toward synchronization of nominally independent 'complementors' and the co-ordination of their externally-held assets. Algorithmic management, big data analytics and internet connectivity permit arms-length control over formally self-employed workers. In this paper, platform capitalism's increasingly widespread use of independent contractors is situated as part of a far more general retreat of states from upholding regulatory frameworks and robust enforcement mechanisms. This encourages the increasing predominance of private law of business contracts and self-regulation in labor markets over public regulation (Lobel, 2004).
In this context we highlight three drivers of platforms' private provision of SEPs. First is the desire to avoid regulatory interference in platform governance, and especially with the contractual status of platform labor (Englert et al., 2021, Cusumano et al., 2021. Second is the fulfilment of the labor reproduction-retention function. And third is the capacity of such agreements to extend the managerial capacities of platforms over their complementorsespecially, but not exclusively, self-employed workers. Collectively, these benefits encourage platforms to (selectively) deliver privatized social and employment protections, under the rubric of contracts for service provision.
Our case study of Hermes' SE+ contract demonstrates both the potential and the limits of platform governance's capacities to replace legally-constructed and state-regulated SEP systems. On the one hand, the form such provisions takeembedded as they are in contract rather than employment law -(intentionally) overlooks the unequal bargaining power of the parties involved. This means they are both patchwork (and do not compare with employee or limb (b) worker statutory SEP provisions), while not being legally guaranteed. It also remains unclear how robust the partnership framework is. The IWGB v Deliveroo case of December 2018 found that trade union recognition agreements for the self-employed are not subject to statutory protection. Should the company elect to unilaterally renegotiate or withdraw the contract from couriers, or decline to continue recognising the union, it seems unlikely that couriers or the GMB would have any legal recourse beyond returning to the courts system to litigate again for worker status. As such, though (unlike elsewhere) no evidence has been found for competition law being used to restrict collective bargaining amongst the self-employed in the UK (see footnote 4 above), the legal foundations of such innovations are nevertheless unstable.
For such reasons, when the Hermes-GMB SE+ and collective bargaining agreements were announced in February 2019, some in the trade union movement announced their trepidation. Jason Moyer-Lee of the IWGB, for instance, complained that the partial coverage of the SE+ contractwhile representing a 'victory for trade union representation'also entailed a threat to employment 'rights' of platform workers due to the 'optional' nature of the contract (quoted in Lo, 2019). But the emergency partnership frameworkdriven in part by the Covid-19 pandemichas strongly incentivized both senior management and regular couriers alongside their union representatives to expand the rollout and enforcement of SE+ model, while substantially strengthening the grassroots union organization recruitment effort. This has ensured considerable improvements in courier renumeration and protections (such as the intervention of the shop stewards to have dismissed couriers' work reinstated), while further contractual enhancements driven by the union's strengthened bargaining position (such as parental leave and pension provision) are said to be objects of future negotiations.
The standard employment contract minimizes labor flexibility and increase its costspotentially spurring firms' technological innovation aimed at saving, rather than squeezing, labor (Aloisi and De Stefano, 2020). But in the context of slow post-2008 economic growth, platforms' use of contract law to engage independent contractors, while provisioning (limited) private social and employment protections, seem likely to represent viable 'good-enough' alternative for governments. States should remain aware, however, both of the patchwork nature of private SEP systems, and the potential threat to their tax bases should such arrangements become widespread. Conversely, for trade unions operating in the virgin territory of the platform economy, (regular) employment status may function more as an ultimatum than a norma stick with which to threaten platforms with litigation in the face of failed negotiations on privatized informal SEP agreements. For these reasons, privatized SEPs appear very likely to grow in scope and significance. Further comparative studies will surely deepen our understanding of how firms and unions are likely to shape platform governance going forwards.

Data Access Statement
Because of the sensitive nature of the research, interviewees did not consent to the retention or sharing of their data.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.