Article body

Rationale

Driven by the promotion of mobile Internet technology and the concept of the sharing economy, platform employment has risen on a large scale around the world. For the purpose of this paper, we define platform employment as work by “people using apps [also commonly known as platforms] to sell their labour” (Wood et al., 2019 : 57). This phenomenon is attracting growing research interest in the nature of employment, its terms and conditions and worker wellbeing, and the implications for labour law and labour relations (e.g., Ashford, Caza and Reid, 2018 ; Cherry and Aloisi, 2017 ; Hall and Krueger, 2018 ; Healy, Nicholson and Pekarek, 2017 ; International Labour Organization, 2021 ; Petriglieri, Ashford and Wrzesniewski, 2019 ; Schor and Attwood-Charles, 2017 ; Stewart and Stanford, 2017). Platform employment began to develop on a large scale with Uber, Lyft and other (US-based) companies that provide online car (taxi) booking. This form of employment has led to a series of legal challenges in different countries to regulate the employment relationship and to provide the worker (service provider) with a level of labour protection.

For example, in Uber BV v Aslam [2021] UKSC 5 (United Kingdom), the British Supreme Court ruled that Uber drivers are “workers” (but not employees), rather than self-employed, and, therefore, qualify for the rights conferred on workers by employment legislation (Bourke, 2021). In March 2020, the French Supreme Court ruled an Uber driver to be an official employee of the platform in the case Arrêt n°374 du 4 mars 2020 (19-13.316) - Cour de cassation - Chambre sociale (France). The main basis of the Supreme Court’s decision is the “subordination relationship” between the platform and the driver : the driver cannot establish contact with the passenger, the work is carried out under the instructions of the platform and the driver cannot alone determine the price. Therefore, the driver should be considered to be an “employee.” However, due to Covid-19, drivers in France are worried that they will be unable to earn income in other fields after becoming Uber “employees,” so they prefer to keep their “self-employed” status. Both the French takeaway driver case in 2018 (Butler, 2018) and the French Uber driver case in 2020 occurred in the absence of clear legislation on the status of platform workers in France. Unlike the Uber case in the UK (Uber BV v Aslam [2021] UKSC 5), in France these precedents do not deviate from the dual framework of salarié – non salarié. The French labour law scholar Isabelle Daugareilh pointed out that the dispute over the legal status of platform workers is an attempt to expand the scope of labour law.[1] However, French jurisprudence has caused some concerns,[2] specifically that it is not conducive to the entry of similar companies into the French market and will further Uber’s monopoly because that company had already established its market dominance by taking advantage of the legal grey area at the time.

The United States 2750.3. California Labor Code is part of the California AB5 Act (An act to amend Section 3351 of, and to add Section 2750.3 to the Labor Code).[3] The Act requires Californian courts to adopt the ABC test instead of the original Borello test mechanism. The test is designed to greatly increase the platform’s burden of proof for denying employment relationships. However, the Act caused great controversy in California, a large number of freelancers protested and the court was forced to introduce a very complex exemption application. The Californian 2020 “Proposition 22” Bill 11.5 was passed by a large margin (58.39% to 41.61%). It confirmed that the ride-hailing driver is legally an independent contractor, not an employee, thus preventing the AB5 Act from being applied to this group. Ride-hailing drivers receive protections that other independent contractors do not enjoy, including wages and other benefits. This marks the addition of a third type of labour to US law : the US legal system has recognized a third type of employment relationship. In the Uber case[4] and the Lyft case in the United States,[5] although the employment relationship was determined in both cases, the judges clearly expressed the dilemma. Both cases were finally settled in higher courts.

Coiquaud and Martin (2020) examined the Ontario Court of Appeal’s decision to invalidate the arbitration clause in Heller v Uber Technologies Inc. There was an imbalance of power between the platform and the workers, and the court intervened to redress the imbalance, albeit far from fully. Moreover, Coiquaud and Martin’s (2020) analysis of “the Ontarian and American decisions regarding the validity of mandatory arbitration agreements between Uber and its drivers” shows “the determining impact of the approach chosen by courts,” suggesting a level of legal uncertainty and judicial flexibility contingent upon the judges.

In short, the above cases in different countries point to the diversity of the legal systems underpinning the court decision. The decision of the Supreme Court of the UK is different from the decision of the Supreme Court of France. The former judged the driver to be a “worker” under the rules of a third type of employment ; the latter judged him to be an “employee” under the “employee–non-employee” dichotomy. Both the decision of the Supreme Court of the UK and the result of the California referendum indicate that flexible employment rights in platform employment can be protected only by creating a third type of employment status. Overall, the above legal examples stress the need to develop a clearer set of laws to guide legal decisions. As Lévesque, Fairbrother and Roby (2020 : 647) argue, “digitalization is disrupting and reordering the regulation of work and employment” and “may lead to organizational and institutional experimentation.”

Mainstream American scholars tend to use the “extent of control” and “economic realities” tests to determine the nature of the (employment) relationship between the platform and the labour provider. The “economic realities” test determines the nature of a business transaction by examining the totality of the commercial circumstances in the US legal context. Courts use this test to ascertain if a person is an employee or an independent contractor. Mainstream American scholars argue that, in most circumstances, the network platform has minimal governance and little control over the labour provider. Thus, an employment relationship cannot be established with either test. In Germany and Japan, subordination theory is applied, and a similar conclusion is reached (e.g., Waas et al., 2017 ; Xie, 2019). Subordination means that the employee provides labour under the employer’s command and supervision. It is the essential characteristic of an employment relationship, being likewise used in Chinese academia and judicial practice to determine the employment relationship (e.g., Feng and Zhang, 2011 ; Li, 2011 ; Xie, 2009 ; Zheng, 2005).

Some scholars have pointed out that subordination is the most prominent feature of the labour contract (cf. Huang, 2003). In many legal systems, subordination is an important indicator or test in determining “if an employment relationship exists” (Davidov, 2017 : 4). Subordination in this context includes personality subordination, economic subordination and organizational subordination (Davidov, 2017 ; Xie, 2019). These central elements often do not exist in platform employment or only partially exist as agency employment, where the employment relationship is weak or defined as independent work because the labour provider is perceived as having a high level of autonomy and independence. This situation invites judicial challenges. In the meantime, “platform capitalism is offering substandard work and increasing inequality” (Schor and Attwood-Charles, 2017 : 11).

With the rise of digital technology, platform employment has emerged almost simultaneously in China and the major developed countries, subsequently becoming the forerunner in terms of business model, service type and job creation. According to the “Sharing Economy Development Report (2021),” there were approximately 84 million service providers in the sharing economy in 2020 (Ministry of Information Industry, 2021). The “Financial Report 2020” of Meituan, China’s largest food delivery platform, revealed that a total of 9.5 million takeaway drivers received income through the Meituan platform in 2020, delivering an average of 27.7 million takeaway orders daily (Meituan, 2021).

Because of the rapid growth in the sharing economy and in platform employment in China, this subject has attracted growing research interest from Chinese scholars (e.g., Wang and Li, 2018 ; Wang and Wang, 2018 ; Xie, 2019 ; Wang, Wu and Yang, 2018 ; Wang and Zhang, 2019 ; Zou, 2017). Their studies focus mainly on legal challenges and practical developments. However, there have been no systematic studies of court decisions on platform/labour provider disputes to shed light on the legal reasoning, the legal dilemma and the broader consequences for society as a whole. Our study fills these research gaps by addressing two research questions :

  1. What is the attitude and role of the courts in resolving disputes between the Internet network platform and the labour provider in China within a civil law system ?

  2. What are the prospects for legal innovations to improve protection of platform labour providers who fall outside the scope of the current labour law, to counter the rapid expansion of digital capitalism at the expense of the under-/unprotected ?

Given the size, speed and variety of the developing platform economy, China offers a rich context for research on platform employment with broader implications for governance of employment relationships and for conceptualization of this new digitally driven form of flexible labour deployment in other national contexts. Platform employment also has significant implications for the evolving role of legal regulation, and its theoretical framework is key to facilitating the transition from organized industrial production within the limits of each firm to organized service delivery within the limits of society, thus making the employment relationship a social relationship. Our study will contribute to emerging practical and academic debates over the future of platform employment, labour relations and social relationships, as well as deliberations on the legislative, judicial and administrative development of platform employment.

Our study draws attention to the social risks of platform employment, with a particular focus on platform drivers and traffic accidents, as the Chinese courts are confronted with the problem of how to protect the rights and interests of third parties (i.e., the public). In the cases of our study, labour service providers accounted for 80% of third-party damage. The damage to third-party rights is a direct reflection of the social risks. Because, in most cases, the courts have ruled out the existence of a formal employment relationship between the platform and the labour provider, the latter must legally bear the financial consequences. This decision may be legitimate but it seems unreasonable and will, in the absence of institutional interventions, have a profound impact on the broader society as highly unregulated platform employment continues to grow and diversify in China. Our study, therefore, provides a strong justification for institutional intervention in the Chinese context, which is relevant to other national settings with expanding platform employment.

We argue that when an economy is largely driven by consumer demand with a strong utilitarian pursuit of cost/profit, efficiency and convenience, the welfare of the workers and the public may be undermined, and that regulatory intervention is needed to avoid a systemic collapse in the pursuit of economic growth. This argument is in line with the concept of “moral economy,” that is, an economy based on social good, fairness and justice that balances economic and social interests (Bolton, Houlihan and Laaser, 2012 ; Götz, 2015). We also call for cross-country comparative studies on this phenomenon, as an example of socially responsible management research, to learn more and to explore legal avenues, both in theory and in practice, and both nationally and internationally. In terms of policy/legislative intervention, we propose adopting a third category of employment status, similar to the one used in the British system, to extend some legal protection to those workers who are neither “employees” nor “self-employed.”

Types of Platform Employment

Currently, China has three categories of platform employment, based on whether or not the platform directly participates in and organizes labour service transactions. In the first type, the platform directly employs labour providers, and the labour contract is clearly defined. This is a traditional employment relationship where the platform is the employer and the labour provider the employee. This employment model is costly because the employer is responsible for the employee’s social security contributions.

In the second type, a specific kind of work is packaged and outsourced to an agency, and the agency organizes with the labour provider to complete the task. The agency signs the labour contract with the labour provider and is legally the “employer,” although the labour provider takes instructions from the platform and bears the platform’s organizational identity when at work. Such arrangements are quite common in other national and industrial settings (e.g., Rubery, Carroll, Cooke, Grugulis and Earnshaw, 2014) and not novel to platform employment.

The third type (Type C hereafter) is the one most commonly adopted in China. The labour providers register and receive orders on a platform app, and can decide whether or not and when and where to provide labour services. They thus enjoy some autonomy. Because they can decide whether they work and the time and venue of work, they are, to a certain extent, freed from the organized production mode of the industrial era and, in principle, have a certain level of flexibility in their mode of labour (see also Wang and Li, 2018 ; Wang et al., 2018). However, the platform can use a reward points system and performance indicators, such as willingness to take orders and customers’ satisfaction/complaints, to shape the provider’s behaviour and elicit cooperation, thus exerting “control” like an employer. Labour providers who receive a certain number of complaint points will be blacklisted and their service no longer used. In this sense, some scholars argue that the platform’s act of blacklisting the labour provider is essentially a “dismissal” (Aloisi, 2016). In addition to these incentive and punitive mechanisms, the labour provider’s earning opportunities and earning level are greatly affected by the platform’s power to determine prices and adjust rates (Cockayne, 2016).

These mechanisms arguably subordinate the labour provider to the platform organization and erode the former’s independence and bargaining power when dealing with the platform and the customer. Labour providers must work at times when they may not wish to work, they must put up with difficult customers and they must accommodate unreasonable customer requests to avoid getting complaints, losing points and possibly being blacklisted. In this sense, their problems are similar to those of service workers (Grandy, Diefendorff and Rupp, 2013). Platform organizations adopt a simple but rather aggressive labour management strategy that denies, on the one hand, the legal employment relationship, and thus any accountability for the labour provider, while, on the other hand, using positive and negative incentives to maximize labour retention and performance and to reduce transactional and reputational costs. Such a strategy is problematic and, amongst other things, creates legal ambiguity in employment status when disputes arise.

In short, the Type C model is flexible in the sense that it fragments work tasks and labour control (Wang and Zhang, 2019 ; Wu and Li, 2018). The labour provider is given autonomy at the price of agreeing to cooperate with the platform and being controlled by the platform. This situation is different from the full independence of the independent contractor ; it is also different from the employer’s command and supervision of the employee in a formal and traditional employment relationship. This form of labour deployment and relationship is a new type of labour transaction mode, which is not (adequately) covered by existing labour law in many other countries (e.g., Cherry and Aloisi, 2017 ; Stewart and Stanford, 2017 ; Xie, 2019). As Stanford (2017 : 385) cogently pointed out : “this ambiguity has so far allowed digital platforms to evade normal obligations imposed on traditional employers – although that immunity is being contested on many fronts.” This form of platform labour utilization has strong implications for reconceptualizing the framework of analysis for employment relations, as well as practical implications for labour protection and social policy and for the broader interests of society.

The rest of our study will focus on the Type C model. While it has the lowest labour cost and risk of all the types of platform organization, it has the highest social risk and is the most contentious one when disputes occur. It also suffers from a high level of uncertainty and unpredictability in court decisions and is prone to attracting public attention, as indicated in the Introduction. In general, the Type C model is an innovative form of employment in the Chinese context. However, key institutional actors, such as the state and the platform organizations, seem to have taken a relaxed attitude toward defining the nature of this form of employment and the associated obligations and entitlements for all parties concerned.

Research Methods

Our study draws primarily on secondary data, namely 102 cases of publicly available court decisions from 2014 to 2019 (data available upon request). There is a strong tradition of using court decision reports and court cases in interdisciplinary socio-legal research to understand legal processes, legal institutions and legal behaviour (e.g., Banakar and Travers, 2005). Recent studies in industrial relations have also started to take advantage of the accessibility made possible by the digitization and publication of court decisions (e.g., Xie, Wang and Cheng, 2017). The case decision reports were downloaded from the Supreme People’s Court “Network of Court Decision Papers,” which are freely available to the public on the Supreme People’s Court Website : http://wenshu.court.gov.cn/. The keyword “platform employment” was used to search for relevant cases without any restriction on time or location. Cases related to the independent mode of employment (see below) were then downloaded for our study. The first case was found in decisions from 2014. This is mainly because three of the five largest and dominant Internet network platforms (all in driving, food delivery and courier services) were established in 2011, 2012, 2013 and 2014 (two of them). For the two founded in 2011 and 2012, they started to grow in size only in 2014, a year that witnessed explosive growth of Chinese platform employment fuelled by aggressive capital investment. As we can see from Table 1, there were relatively few cases in 2014-2016. The lead author searched, downloaded and analyzed all the reports and categorized them into tables. Case reports were analyzed by means of manual content analysis (Miles and Huberman, 1994).

Main Disputes and Judicial Differences in Internet Platform Employment in China

Although platform employment has become a universal and large-scale mode of work (e.g., Wood et al., 2019), China’s existing legal system has not yet provided clear regulation. Existing regulation appears to avoid the problem of defining the nature of the relationship between the platform and the labour provider. Under the current Chinese legal framework, the types of contracts subject to labour transactions are divided into two categories : labour contracts and civil contracts. The fundamental difference is that the labour contract is covered by a large number of mandatory regulations. Labour law provides workers with a comprehensive, systematic level of protection, and the employer is responsible for the employee’s duties and conduct. Civil contracts, by contrast, do not have such mandatory provisions, and labour providers are usually responsible for their actions.

Under a labour contract, the employee is seen as working under the employer’s direction with no independence and as “subordinate labour.” Under a civil contract, labour users and providers are considered to be equal. The main party in the contract is “independent labour.” Consequently, the legislative framework is a dual structure of “subordinate labour-independent labour.” Because, for the platform, the mandatory provisions of a labour contract mean higher costs, the rational choice is to enter into a civil contract with the labour provider. With such a contract, the platform is in the stronger position, the labour provider being an individual with little if any bargaining power.

The story does not end there. Once a dispute arises with the platform, the labour provider can initiate a lawsuit and appeal to the court to re-judge the nature of the relationship between the platform and the labour provider. The court can also rule on the sharing of responsibilities. At this point, court decisions suffer from legislative evasion of the need to define the nature of the relationship, the result being judicial differences and discrepancies.

In the Supreme People’s Court “Network of Court Decision Papers,” for platform employment dispute decisions during 2014-2019, we retrieved 102 cases that concerned the Type C model. The cases mainly involved driving, food delivery and courier services. All of the disputes involved road accidents. As shown in Table 1, dispute cases rose sharply, peaked in 2017, started to drop in 2018 and fell even more in 2019. The disputes seem to have educated people on both sides, with the result that more precautions are being taken.

Table 1

Distribution of Dispute Cases Relating to Labour Services in the Platform C Model by Year

Distribution of Dispute Cases Relating to Labour Services in the Platform C Model by Year

Note : The year of decision is the year when the court accepted the case. Because the court needs a certain length of time to make a decision, the acceptance of a case and the decision on the case may not always be in the same year.

-> See the list of tables

Disputes mainly occurred in cities with the most developed platforms and a Type C model of employment. Over half the cases were in Beijing and Shanghai (32% and 26% of the total respectively), a reflection of the economic affluence and importance of the two cities. With the expansion of platform employment, such disputes have also started to occur in other major cities.

The disputes generally followed a simple plot, and the evidence was straightforward. The platform and the labour provider had little disagreement over the facts, and the dispute usually focused on the nature of the relationship between the two parties. Two-thirds of the cases were terminated after the first trial, and the parties accepted the court’s decision. About one-third of the cases went to a second trial. At the second trial, the court maintained a rather conservative stance : out of 30 second-trial cases, only 10% of the decisions were changed. In the absence of a statutory basis and judicial consensus, it is clear that the courts tend to respect the decision of the first trial in determining the relationship between the platform and the labour provider.

Current disputes fall into two main categories. In one category, the labour provider asks the court to determine that a labour relationship exists with the platform. Typically, the labour provider had been injured while providing the service and now hoped to receive insurance benefits for a work-related injury by proving an employment relationship with the platform. In the other category, the labour provider caused damage to a third party while providing the service. In the case of liability for damages, the court must judge the contractual relationship between the platform and the labour provider. The legal loopholes are precisely where the judiciary has a decisive influence on the nature of the relationship. If the court determines that the relationship is a labour contract, the employer shall be liable for the labour provider’s injury and for the damage caused by the labour provider to the third party.

If, however, the court determines that it is a civil relationship, the labour provider shall be personally liable. As shown in Table 2, the majority of the cases (80%) involved third-party damages. It seems that the labour provider was seeking to offload financial responsibility to the platform by establishing an employment relationship with it and thus making it legally liable for the third-party injury. The motive was purely financial, as such a burden is prohibitively costly for labour providers, many of whom are already financially hard up, and who would be plunged into serious debt and lifelong poverty with their families if they honoured their liability to the injured third party. Turning to the platform was essentially their last resort.

Table 2

Distribution of Cases by Reason for Dispute

Distribution of Cases by Reason for Dispute

-> See the list of tables

If the labour provider is not injured and only wants the court to confirm the labour relationship, the court will generally turn down the request. If the labour provider is injured and has an urgent medical need, there is a greater chance of obtaining court support. One reason is that the court has much leeway in the absence of explicit legal instructions, and will often expand its interpretation of labour relations on the basis of a humanitarian stance to support the weak and the needy. This is especially true in situations where the court is only supporting a compensation claim for work-related injuries. The platform, as a powerful economic entity, has sufficient capital to pay the compensation and spare the injured labour provider and his/her family the financial hardship (healthcare is prohibitively expensive in China even with medical insurance). Such a decision will achieve a good social outcome. However, as Table 3 shows, the court did not support the worker’s request in the majority of cases.

It is worth noting that the platform usually purchases personal insurance on behalf of the labour provider out of the charges (three yuan per day) it deducts from his/her pay. An amount of 100,000 yuan (approx. US$15,000) can cover most minor injuries sufficiently but not more serious ones like bone fractures, which are quite common in traffic accidents. However, following a driver’s sudden death in December 2020 and the small compensation that the platform initially paid out to his family, it was revealed that the platform retained a large proportion of the insurance fee (1.94 yuan) it had deducted and thus made a profit from his death. The government ordered that all insurance fees deducted from labour providers be fully used for the insurance contribution. In a subsequent review, the government found that labour providers could actually pay much less than three yuan to receive reasonable insurance coverage. Reform in this area is pending (Red Star News, 8 June 2021).

Table 3

Court Decision Outcomes in Cases of a Request for Recognition of a Labour Relationship

Court Decision Outcomes in Cases of a Request for Recognition of a Labour Relationship

-> See the list of tables

Because platform employment in China most often involves transportation, such as driving, food delivery and courier services, most injuries are due to traffic accidents. In addition to self-inflicted damage, a large number of the cases involved injuries/damages caused by the labour provider to a third party. This third party can be understood as the public. To be more cost-effective, labour providers often violate traffic and platform regulations, and the platform often turns a blind eye to such behaviour.

The existing court decisions suggest that the court did not hold a clear position on how to distribute this social risk. As Table 4 indicates, the labour provider was found to be responsible for more than half (64%) the cases. However, in around 35% of the cases, the court found the platform to be responsible, even though the situations were similar. The intention was to protect labour by assigning legal responsibility to the platform as the “employer.” Such legal reasoning is pragmatic, result-driven and heavily influenced by the judge’s personal motivation, rather than by judicial caution and consistency. As the number of disputes associated with platform employment grows, such an approach will become unsustainable and undesirable if the country is to maintain and enhance its economic growth, social stability and judicial competence and reputation.

To assign the platform responsibility is to expand the interpretation of the labour relationship. Despite insufficient legal logic, quite a few courts have ruled in this way to achieve better social effects and allocate responsibility for social risks. In one particular case, the court ruled that the labour provider and the platform should share the liability. The court concluded that this was a new kind of labour relationship (i.e., different from the traditional employment relationship) and that the responsibility (cost) should be shared by both parties. The courts are therefore faced with a dilemma : the current law is inadequate for regulating platform employment and the court is hampered in making decisions by legislative, pragmatic and cognitive ambiguity and constraints.

Table 4

Decision Outcomes of Cases in Which the Labour Provider Has Caused Damage to a Third Party

Decision Outcomes of Cases in Which the Labour Provider Has Caused Damage to a Third Party

-> See the list of tables

In sum, in the absence of clear regulations, the courts have reached two opposite conclusions, although in most cases they have ruled that an employment relationship does not exist between the platform and the labour provider. To some extent, these opposing decisions, based on largely similar facts across the cases, reveal the logical pathway leading to the court decisions depicted in Figure 1.

Figure 1

Pathway of Outcomes for Court Decisions on Platform Employment

Pathway of Outcomes for Court Decisions on Platform Employment

-> See the list of figures

The main differences in the decisions can be explained by two main questions and how they are resolved. One question is whether the network platform “has used labour service” from the labour provider. If the platform is judged to be an intermediary (i.e., in a non-employment relationship), its behaviour is viewed as being one of only providing information. The other question is whether the network platform has “management control” over the labour provider. If the platform is judged to have used the labour service, the next question is whether the platform imposes management control over the labour provider while “using the labour service.” The reason is that a labour relationship must meet two criteria : “use of labour service” and “management control.” As to what “management control” means, the meaning is drawn from subordination theory (Davidov, 2017 ; Xie, 2019). According to subordination theory, the court must consider the following : whether the platform provides the labour provider with training ; whether the platform’s rules and regulations apply to the labour provider ; how the platform pays the labour provider ; and whether the platform’s business is served by the labour provider (Wang, 2016).

Discussion

Two issues stemming from the above analysis warrant further discussion : the limited utility of current labour law ; and how to conceptualize and thus regulate the platform’s supply of on-demand services to customers.

Obstacles to Applying Labour Law to Platform Work

As alluded to earlier, existing labour law is inadequate in dealing with certain types of platform employment. The labour regulation system is based on subordination theory, that is, on the dependence of labour. Such subordination is often lacking in the Type C model. Based on subordination theory, labour law cannot cover network platform labour providers because they are insufficiently subordinated. The inadequacy lies in the boundary defined by subordination theory. Here, labour providers challenge the use of subordination theory in labour law in two ways.

The first challenge is conceptual. If labour law remains based on subordination theory, it will increasingly be inadequate in regulating new forms of employment and work relationship. Although China’s current production methods are already diverse, its judiciary still generally relies on subordination theory to establish a labour relationship. There is therefore scope and necessity to improve/amend the system norms that are based on subordination theory. Here, German legal practices may be a good example. In Germany, “employment at will” is not permitted. Although the production mode has changed, the nature of the labour service under the employer’s supervision has not, and therefore subordination theory continues to apply (Wang and Zhu, 2015).

The second challenge is substantive. Subordination theory recognizes that personal subordination is key to the employment relationship. Despite the weakening of this feature in platform employment, there is still economic subordination (i.e., platform income is the labour provider’s main source of livelihood). A technical solution is to go beyond personal subordination and focus on economic subordination as the main basis, thus expanding current labour law’s scope of application. However, this is not a sustainable solution because it essentially expands the scope of labour law by challenging and amending subordination theory. It does not provide independent platform labour with proper legal protection, sometimes with negative consequences for the customers they serve and the general public, such as in road accidents. A long-term and more effective solution is to develop a relevant regulatory mechanism based on the characteristics of the labour service and apply it to the new form of employment relationship.

Platform Employment and the Socialization of Contract Service

The Type C model is innovative in the Chinese context in that it collects scattered information about demand for labour services. The platform serves as a unified service provider, which channels this scattered information about a large volume of market demand. It is this market demand that supports the business model of organizational platforms, which has turned platform employment into a burgeoning independent industry in China, where the majority of employment has already become informalized (Cooke et al., 2019).

In the Type C model, becoming a labour provider requires only a simple registration and identity authentication procedure. There are no requirements, such as skills or age, nor are working hours and locations stipulated. Therefore, a large number of (rural) migrant workers have left their jobs in construction, manufacturing and general services to participate in platform employment full-time. This growing segment of the labour market is also being joined by people who are already in full-time employment but wish to work on the platform in their spare time for various reasons.

As there is no restriction on how many platforms one can register with, a labour provider may register with several and pick and choose job orders according to his/her preference, to maximize time efficiency. In this context, labour providers and service requesters have become large and expanding social groups that fuel labour demand and supply via a mobile APP connected to the platform. Their relationship is transactional. As such, the legal dilemma stems from the network of individual platforms that provide large-scale socialized services, for labour law alone does not hold sufficient power to regulate the large and spontaneous movement of labour demand and supply, within which diverse and innovative businesses are frequently emerging. In fact, Stanford (2017 : 397) argued that the “‘social licence’ of digital platform businesses would be vulnerable in the event that the impacts of their labour practices become the focus for public attention and concern,” and called for a rejection of the model of precarious work adopted by most digital platforms.

Legislative Orientation and Regulatory Approach toward Socialized Platform Employment

To regulate platform employment, the point of departure, then, is to socialize labour deployment, to break away from the dual legislative framework of “subordinate labour-independent labour” and build a multi-level security network. Many countries, such as the US, France and Japan, have constructed a regulatory system based on the employee/self-employed dichotomy, which rests on the dichotomy of “independent labour-dependent labour” (e.g., Waas et al., 2017). This type of legislative model will recognize only “one or the other” and is in line with the institutional logic that labour law should protect dependent labourers, and not other types of labour providers.

Another legislative model is to regard independent labour-dependent labour as two poles with other types in between and provide these in-between workers with a certain level of legal protection. For example, the scope of “worker” in British law is broader than that of “employee,” and those who are deemed “workers” instead of “self-employed” are entitled to some labour protection. In the case of China, the labour law covers only employees. There is no “worker” category that is entitled to a level of protection as exists under British labour law. Hence the Chinese government can consider adopting this approach, although a worker category may also create its own legal challenges as companies continue to find ways to bypass legal constraints (Cherry and Aloisi, 2017).

Regardless of the legislative model, there is limited scope for adjustment of the existing labour law. Platform workers in China without “employee” status can obtain cross-platform employment, can autonomously control working hours and are extremely scattered and difficult to organize. These characteristics make it difficult for them to benefit from the existing working hour standards, collective wage negotiation and work-related injury insurance stipulated in the labour law. Moreover, given the reality that China’s labour law system is under-developed, it would be better to develop platform employment standards and protect the rights and interests of labour providers outside traditional labour relations by focusing on defining new systems, rather than stretching current definitions to bring platform workers under existing labour laws.

Therefore, the real legal problem for network platform labour providers in China is how to provide them (and the public) with appropriate protection mechanisms, rather than contemplating how they can be included in the current labour law. Given time, the risks associated with individual workers, as borne out in the dispute cases (i.e., unemployment due to injury, poverty in old age, poverty due to lack of social security protection), will aggregate into a substantial burden of social risks that cannot be easily written off. Legal efforts should be directed to exploring the establishment of a multi-level legal protection network. Possibly relevant here may be Stewart and Stanford’s (2017) proposed options for extending regulation to accommodate gig work in the Australian context. In particular, they suggest confirming and enforcing existing laws, clarifying or expanding definitions of employment, creating a new category of “independent worker,” creating rights for workers, not employees, and reconsidering the concept of “employer” (Stewart and Stanford, 2017).

In addition, to protect the rights and interests of labour providers, we propose focusing on their income and occupational health and safety protection by carrying out the following four main actions. Our argument is informed by the concept of moral economy : social good, fairness and justice should be brought into balance with the economic and social interests of different stakeholders in society as a whole for sustainable development (Bolton et al., 2012 ; Götz, 2015). First, there should be a pricing and remuneration guarantee system that is independent of the platform organizations. This mechanism would offer some protection of the labour provider’s right to survive. A tripartite system, consisting of representatives of the local labour authority, platform organizations and trade union organizations, can be set up to determine pricing levels and subsequent price adjustments. Second, to prevent driver fatigue, there should be a monitoring and control system to regulate the length of time that the labour provider is continuously online. A platform employment works council (PEWC) can be set up to determine pricing, working time and other related issues. Third, an occupational risk protection system should be put in place to provide affordable work-injury social security protection for long-term treatment and rehabilitation. Fourth, an appeal and relief system for disputes should be set up to provide platform labour providers with an external mechanism to file their complaints and grievances and to seek remedy. Such a mediation function can be assumed by the PEWC. If dissatisfied with the PEWC’s proposed resolution, the labour provider can appeal to the court.

In making this proposal we seek to address the problems we have identified through our analysis of platform employment and dispute cases in particular. We acknowledge that our suggestions may be challenging to adopt, as platforms are powerful and some, in collusion with other key stakeholders, are even deliberately destroying/bypassing current labour regulation systems or preventing them from being developed. Nevertheless, in a country like China, where administrative policy can be developed and implemented quite rapidly to address significant socio-economic problems, there may be opportunities for the government to act if there is sufficient pressure. The promulgation of the “Opinions of the General Office of the State Council on Supporting Multi-channel Flexible Employment” (Guobanfa [2020] No. 27) issued on 28 July 2020 could be seen as a step in this direction, although its impact remains to be seen.

Conclusions

Platform employment in China is a substantial and growing segment of informal employment that operates in a highly unregulated and competitive market. Academic debate is emerging on the subject but remains unconsolidated and often makes no differentiation between the varieties of platform employment and their respective implications for labour protection and regulation. Our study makes two related contributions.

First, we examine the attitude and role of courts in judging disputes between platforms and labour providers, within legal constraints. Their approach is conservative, pragmatic and yet sympathetic toward labour in some cases, where such leniency makes a significant impact. A level of judicial flexibility does exist, but the exercise of such discretion in favour of labour is contingent upon, amongst other things, the judges’ ideological values. While our study reveals the judges’ critical role as institutional actors at the micro level, the protection of millions of workers in China’s informalized labour market, where labour law enforcement is ineffective (e.g., Cooney, 2007), cannot be contingent upon the goodwill and moral conscience of a few judges. Instead, legal provisions should be introduced to institutionalize protection that goes beyond the current labour law both conceptually and substantively.

That contribution leads to our second one. We propose that a socialization of contract service should be central to platform employment. Due to the low entry barrier to employment for platform agency firms (i.e., firms that use a platform to operate their businesses, such as online takeaway services), and the lack of protection and regulation of seemingly independent labour providers, this form of platform employment has incurred additional risks to the individual and society at large. The efficacy of existing labour law, even when allowing for the courts’ most liberal interpretation of the contractual relationship between the platform and the labour provider, has proven to be limited in regulating this type of relationship and addressing practical needs. To this end, we call for a departure from the subordination orientation that underpins traditional labour law when contemplating a regulatory framework for those engaged in the Type C model. Instead, a multi-level platform employment security network should be developed to provide a better balance between economic efficiency and social justice.

Our study has a number of practical implications, particularly for regulatory bodies and policy making. One suggestion is to register platforms in terms of a particular set of classifications. This will help align legal support with targeted protections and sanctions. Here, the requirements of the Labour Contract Law for labour dispatch companies may be relevant and adapted to regulate the entry of platform agency firms. Another suggestion is to maintain labour standards by raising the entry standard for platform agency firms and creating an entry barrier. Current policy focuses mainly on the technical and innovative aspects of platform employment. Large firms usually enter the industry with a direct employment model. Once they become established and are profitable, they tend to adopt the subcontracting model to detach themselves from a formal employment relationship. Because the requirements of platform sub-contractors are unregulated, platforms are free to abuse their dominant position. Policy may be developed to specify qualifications for platform employment agencies. When platform labour providers sign contracts with sub-contractors, the latter should be regarded as their employees, or the law could be developed to afford them the ‘worker’ status that exists under British labour law, thus bringing in some form of formal employment relationship with associated obligations and entitlements on both sides. Here, a useful framework and reference point may be the 1997 “Private Employment Agencies Convention” (PREA Convention) of the International Labour Organization (ILO, C181). It not only regulates the platform agency firms that employ workers directly but also unmasks the platform by bringing it to the fore and holding it accountable for labour standards via platform agency firms (see Wouters, 2018). As De Stefano (2016 : 123) argued, labour problems associated with the gig economy should not be treated in isolation. Instead, they should be viewed as part of the broader trend of “demutualization of risks, and informalization of the formal economy.” As such, they should be included when contemplating “comprehensive solutions to labour problems in modern and future labour markets.”

Our study has some limitations. First, we examined secondary data from a legal perspective. We thus provide only a snapshot of the situation of platform employment and dispute outcomes. In the future, researchers may adopt a mixed-method approach to examine the employment relations of gig workers more broadly and deeply, including work intensity, incentives and penalties, nature of work, forms of coping and resistance, and well-being. Second, we focused mainly on the low-skilled and low-paid segment of platform employment, where the workers are most in need of protection. Future studies may examine the employment situations of those in the relatively high-skilled and well-remunerated segment of the platform economy. In sum, by studying the work, employment and labour relations of platform employment as a form of gig work in the Chinese context, it is possible to gain useful data for cross-country comparison as well as lessons for other countries. We believe that the existing analytical framework of industrial relations should be reconceptualized, or broadened, to include non-traditional types of employment and non-traditional institutional actors with a view to regulating employment relations not necessarily within the workplace but wherever work takes place, not only between employers and employees but also between businesses and workers.